US9003053B2 - Message acceleration - Google Patents

Message acceleration Download PDF

Info

Publication number
US9003053B2
US9003053B2 US13/624,788 US201213624788A US9003053B2 US 9003053 B2 US9003053 B2 US 9003053B2 US 201213624788 A US201213624788 A US 201213624788A US 9003053 B2 US9003053 B2 US 9003053B2
Authority
US
United States
Prior art keywords
data
upper layer
processing system
message engine
protocol stack
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active, expires
Application number
US13/624,788
Other versions
US20130080651A1 (en
Inventor
Steven L. Pope
David J. Riddoch
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Solarflare Communications Inc
Original Assignee
Solarflare Communications Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to US201161538055P priority Critical
Application filed by Solarflare Communications Inc filed Critical Solarflare Communications Inc
Priority to US13/624,788 priority patent/US9003053B2/en
Assigned to SOLARFLARE COMMUNICATIONS, INC. reassignment SOLARFLARE COMMUNICATIONS, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: RIDDOCH, DAVID J., POPE, STEVEN L.
Assigned to SOLARFLARE COMMUNICATIONS INC. reassignment SOLARFLARE COMMUNICATIONS INC. CHANGE OF ADDRESS OF THE ASSIGNEE Assignors: SOLARFLARE COMMUNICATIONS INC.
Priority claimed from US13/754,792 external-priority patent/US20130145035A1/en
Publication of US20130080651A1 publication Critical patent/US20130080651A1/en
Priority claimed from US14/231,510 external-priority patent/US9674318B2/en
Publication of US9003053B2 publication Critical patent/US9003053B2/en
Application granted granted Critical
Assigned to PACIFIC WESTERN BANK reassignment PACIFIC WESTERN BANK SECURITY INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SOLARFLARE COMMUNICATIONS, INC.
Assigned to ALLY BANK reassignment ALLY BANK SECURITY INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SOLARFLARE COMMUNICATIONS, INC.
Assigned to SOLARFLARE COMMUNICATIONS, INC. reassignment SOLARFLARE COMMUNICATIONS, INC. RELEASE BY SECURED PARTY (SEE DOCUMENT FOR DETAILS). Assignors: PACIFIC WESTERN BANK
Application status is Active legal-status Critical
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00Application independent communication protocol aspects or techniques in packet data networks
    • H04L69/16Transmission control protocol/internet protocol [TCP/IP] or user datagram protocol [UDP]
    • H04L69/161Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00Application independent communication protocol aspects or techniques in packet data networks
    • H04L69/12Protocol engines, e.g. VLSIs or transputers
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00Application independent communication protocol aspects or techniques in packet data networks
    • H04L69/30Definitions, standards or architectural aspects of layered protocol stacks
    • H04L69/32High level architectural aspects of 7-layer open systems interconnection [OSI] type protocol stacks

Abstract

A data processing system comprising: a host computing device supporting an operating system and a network protocol stack, the network protocol stack being operable to support one or more transport streams by performing transport stream protocol processing of data packets received over the streams; a network interface device arranged to couple the host computing device to a network and operable to receive data packets over a transport stream supported by the network protocol stack; and a message engine configured to perform upper layer protocol processing; wherein the network interface device is configured to, on receiving a data packet over one of a predetermined set of transport streams, pass the payload data of the data packet to the message engine and the message engine is configured to, in response to receiving the payload data, identify and process any upper layer messages in the payload data in accordance with the upper layer protocol.

Description

BACKGROUND OF THE INVENTION

This invention relates to a data processing system comprising a host computing device configured to perform transport protocol processing and a network interface device configured to perform upper layer protocol processing so as to support low latency upper layer messaging.

In typical computer systems, streaming protocols such as TCP are generally handled in software at a network protocol stack supported at the computer system. This is because handling streaming protocols is expensive to implement in hardware (e.g. at a network interface device) in terms of both processing power and memory requirements. For example, data is transmitted over TCP by encapsulating it in TCP segments that could be dropped, retransmitted and reordered, and as a result a stream of data may not arrive at the receiver in the correct sequence. Thus, TCP streams require that the protocol stack managing the streams can handle out of order packets (which requires a significant amount of storage) and both the receive and transmit paths of a given stream (because the receive and transmit paths of a TCP stream are not independent).

Nevertheless, due to the potential improvements in latency offered by processing TCP streams at a NIC, network interface devices that can perform TCP processing in hardware at a TCP offload engine (TOE) have been developed. Examples include NICs based on the Chelsio T4 and Broadcom BCM5708C chips. Performing protocol processing at the NIC also opens up the possibility of performing upper layer protocol processing at the NIC so as to further minimise the latency associated with communicating upper layer messages over TCP. This is because the transport layer TCP protocol must generally be processed prior to any upper layer protocols—for example, TCP data packets received at a NIC must be correctly reassembled into an ordered stream so as to allow the payload data of those data packets to be re-formed and the upper layer messages extracted for processing by the upper layer protocol stack. Thus, if TCP processing is performed at a NIC, the upper layer protocol processing can also be moved into hardware at the NIC, and the latency associated with upper layer messaging can potentially be reduced.

However, incorporating both a TCP offload engine and an upper layer protocol accelerator in hardware at a network interface device requires even greater processing power and memory to be present at the NIC. This significantly increases the complexity and cost of the network interface device, especially since bespoke hardware solutions such as FPGAs (field programmable gate arrays) are typically used to provide upper layer message processing specific to a particular application. Furthermore, the complexity of an integrated hardware TCP and upper layer protocol solution generally results in long development times and, since the hardware implementing the TCP offload engine and accelerator cannot generally be reconfigured with the same ease with which an updated software transport library can be installed into a host computer system, it is much more difficult to make changes to a hardware implementation (e.g. to fix bugs).

There is therefore a need for a host computer system and network interface device that can support low latency messaging over an upper layer protocol without requiring the complexity and expense of a TCP offload engine in hardware.

SUMMARY

According to a first aspect of the present invention there is provided a data processing system comprising: a host computing device supporting an operating system and a network protocol stack, the network protocol stack being operable to support one or more transport streams by performing transport stream protocol processing of data packets received over the streams; a network interface device arranged to couple the host computing device to a network and operable to receive data packets over a transport stream supported by the network protocol stack; and a message engine configured to perform upper layer protocol processing; wherein the network interface device is configured to, on receiving a data packet over one of a predetermined set of transport streams, pass the payload data of the data packet to the message engine and the message engine is configured to, in response to receiving the payload data, identify and process any upper layer messages in the payload data in accordance with the upper layer protocol.

Preferably the network interface device or message engine is further configured to pass at least the headers of the data packet to the network protocol stack so as to cause the network protocol stack to perform transport protocol processing of the received data packet.

Preferably the network protocol stack is further configured to, in response to receiving at least the headers of the data packet, perform lower layer protocol processing of the received data packet so as to complete protocol processing of network protocols below the transport layer.

Preferably the network interface device passes the payload data to the message engine by making available at least the payload data to the message engine, and optionally the received data packet including its headers and the payload data.

Preferably the network interface device or message engine parses the received data packet so as to identify its payload data.

Preferably the payload data does not comprise any data relating to intervening protocol layers between the upper layer protocol and the transport stream protocol. Preferably the payload data only comprises data relating to the upper layer protocol.

Preferably the message engine is configured to, if the upper layer processing is successful, pass data generated by the upper layer processing to the upper layer endpoint to which the upper layer messages are directed.

Preferably the message engine passes data generated by the upper layer processing to the network protocol stack by writing the data generated by the upper layer processing to a buffer at the host computing device that is accessible to the network protocol stack. Suitably the message engine is configured to, if the upper layer processing is successful, pass data generated by the upper layer processing to the network protocol stack and the network protocol stack is configured to release the generated data to the upper layer endpoint to which the upper layer messages are directed only if the transport stream processing of the data packet is successful and the data packet is determined to have been received in-sequence.

Preferably the network protocol stack is configured to, if the network protocol stack determines that the data packet was received out-of-sequence, discard the data generated by the message engine and cause the upper layer protocol processing of the data packet to be performed in software at the host computing device.

Preferably the message engine is configured to, if the upper layer processing is unsuccessful, signal the network protocol stack to indicate that upper layer processing was unsuccessful, and the network protocol stack is configured to, in response to the signal, cause the upper layer protocol processing of the data packet to be performed in software at the host computing device. Suitably the upper layer protocol processing is performed in software at the network protocol stack or at an application supported by the operating system of the host computing device.

Preferably the message engine is further configured to maintain a packet sequence number for each of the predetermined set of transport streams and to, if the payload data comprises a partial upper layer message, perform the following steps: process the partial upper layer message in accordance with the upper layer protocol; store metadata describing the processing state of the message engine; and on the network interface device receiving a subsequent data packet over the one of the predetermined set of transport streams, check that the packet sequence number of the subsequent data packet indicates that it is the next data packet in the sequence of data packets transmitted over that transport stream, and if so: load the metadata describing the processing state of the message engine and continue upper layer protocol processing of the respective upper layer message; and otherwise, abandon the upper layer protocol processing of the respective upper layer message.

Preferably the message engine is further configured to, on upper layer protocol processing of the respective upper layer message being completed, pass data generated by the upper layer processing to the endpoint to which the upper layer messages are directed.

Preferably the message engine is further configured to, if upper layer protocol processing of the respective upper layer message is abandoned, signal to the network protocol stack so as to cause the upper layer protocol processing of the respective data packets to be performed in software at the host computing device.

Suitably the upper layer endpoint is an application supported by the operating system of the host computing device, the message engine itself, or another hardware engine of the data processing system. Suitably the another hardware engine is a financial trading engine provided at a peripheral device of the host computing system and the upper layer protocol is FIX.

According to a second aspect of the present invention there is provided a data processing system comprising: a host computing device supporting an operating system and a network protocol stack, the network protocol stack being operable to support one or more transport streams by performing transport stream protocol processing of data packets received over the streams; a network interface device arranged to couple the host computing device to a network and operable to transmit data packets over a transport stream supported by the network protocol stack; and a message engine configured to perform upper layer protocol processing; wherein the network protocol stack is configured to, on upper layer data being for transmission over one of a predetermined set of transport streams, form transport stream information sufficient for the message engine to form a data packet without performing transport stream protocol processing and cause the message engine to process the upper layer data in accordance with the upper layer protocol so as to form one or more upper layer messages; and the message engine being configured to, in dependence on the transport stream information, encapsulate the one or more upper layer messages in one or more data packets and cause the network interface device to transmit the data packet over the transport stream.

Suitably the transport stream protocol is TCP and the transport stream information includes: segment sequence number, TCP receive window size, ACK number, source port number, destination port number, maximum segment size, and the TCP options of the TCP stream.

Suitably the upper layer data for transmission is generated at the message engine, at an application supported at the operating system of the host computing device, or at another hardware engine of the data processing system. Suitably the another hardware engine is a financial trading engine provided at a peripheral device of the host computing system and the upper layer protocol is FIX.

Preferably the transport stream information comprises information sufficient for the message engine to form the data packet without performing lower layer protocol processing. Suitably the transport stream information includes one or more of an Ethernet MAC address, a VLAN-id, an IP address and an IP ID.

Suitably the transport stream information comprises a data packet template.

Preferably the message engine is configured to request transport stream information from the network protocol stack on the message engine receiving upper layer data for transmission.

Preferably the transport stream information includes a packet sequence number and the message engine is configured to, if the one or more upper layer messages require more than one data packet, for each additional data packet increment the packet sequence number.

Preferably the message engine is configured to make available to the network protocol stack each data packet it causes to be transmitted.

Preferably the network protocol stack is configured to, if a data packet is to be retransmitted over one of the predetermined set of transport streams, effect retransmission of that data packet by modifying a copy of the data packet made available to it by the message engine and causing the modified copy to be transmitted by the network interface device.

Preferably the message engine is configured to signal to the network protocol stack the packet sequence numbers of any additional data packets it causes to be transmitted.

Preferably the network protocol stack is further configured to signal to the message engine the amount of payload data the message engine is permitted to cause to be transmitted by the network interface device. Preferably the amount of payload data is expressed as a number of data packets or a length of payload data.

Suitably the network protocol stack has access to the upper layer data for transmission and is configured to, if a data packet is to be retransmitted over one of the predetermined set of transport streams, effect retransmission of that data packet by: passing the respective upper layer data for transmission to the message engine along with transport stream information sufficient to generate the data packet for retransmission; or passing the respective upper layer data for transmission to a software entity of the host computing device so as to cause that entity to form the respective one or more upper layer messages and forming the data packet for retransmission at the network protocol stack.

Suitably the lower layer protocols include one or more of IP, IPSec, IEEE 802.3, and IEEE 802.11.

Preferably the network interface device is not operable to perform transport stream protocol processing to completion.

Preferably the message engine is provided at the network interface device or at another peripheral device of the data processing system.

Preferably the network protocol stack is a user level network protocol stack.

Suitably the transport stream protocol is TCP.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be described by way of example with reference to the accompanying drawings, in which:

FIG. 1 is a schematic diagram of a data processing system configured in accordance with the present invention.

FIG. 2 illustrates the structure of a transport stream data packet carrying two upper layer protocol messages.

FIG. 3 illustrates two transport stream data packets together carrying three upper layer protocol messages.

DETAILED DESCRIPTION

The following description is presented to enable any person skilled in the art to make and use the invention, and is provided in the context of a particular application. Various modifications to the disclosed embodiments will be readily apparent to those skilled in the art.

The general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

The present invention relates to a host computing device and network interface device that together comprise a data processing system. A host computing device could be any kind of computing device, including: a server, a personal computer, a switch, a router, a bridge, a virtualised system (e.g. a virtualised operating system and the applications it supports), and a portable device such as a tablet computer or smartphone. A network interface device could be any hardware device configured to interface a wired or wireless network to a host computing device. The network interface device could be provided in any manner between a network and host device, including: as one or more peripheral devices (such as a PCIe card) of the host computing device, and/or as an integral part of the host computing device (e.g. a LAN on mainboard, or LOM chipset).

A data processing system configured in accordance with a first aspect of the present invention is shown in FIG. 1. The data processing system 100 comprises a host computing device 101 coupled to a network interface device 102 that is arranged to interface the host to network 103. The host computing device includes an operating system 104 supporting one or more applications 106 and a network protocol stack 105. The network protocol stack could be provided at the operating system and operate in the kernel context, or it could be a user level protocol stack. The operating system could be a conventional monolithic operating system, a privileged domain (e.g. a hypervisor) supporting one or more virtualised operating systems, or a virtualised operating system itself. There could therefore be multiple software environments (represented by the entities 104, 105 and 106) at the host computing device.

The network protocol stack is configured to process messages formed at the one or more applications 106 so as to encapsulate each message in one or more data packets in accordance with a transport stream protocol such as TCP. The protocol stack would typically also perform protocol processing according to one or more additional protocols. For example, if NIC 102 is an Ethernet NIC, the stack could be configured to form data packets in accordance with TCP, network layer internet protocol (IP) and link layer Ethernet (IEEE 802.3) so as to form data packets having a nested set of protocol headers as follows:

[Ethernet[IP[TCP<payload data>]]]

where transport layer TCP is carried within an IP packet which is carried within an Ethernet packet. It is well known in the art to perform link (e.g. Ethernet) and network layer (e.g. IP) protocol processing at either a host device or a NIC and this will not be discussed further. There may also be additional protocol layers above the transport layer—for example, quality of service protocol layers—and these can also be handled according to any suitable manner in the art, but would preferably be handled at a host device. The present invention concerns a transport stream protocol such as TCP and its payload data representing upper layer protocol messages.

The network protocol stack is shown in FIG. 1 as a single homogeneous unit but could in fact be a collection of multiple processing entities located in hardware (e.g. at the NIC), and/or at the operating system (e.g. running in the kernel context), and/or at user level (e.g. at a user level transport library). The network protocol stack could additionally be capable of handling data flows in accordance with any number of other connected-oriented or connectionless protocols.

The host computing device may transmit data over network 103 by means of the network protocol stack by passing messages to the stack for transmission in accordance with the transport stream protocol. Typically those messages will be configured in accordance with an upper layer protocol such as HTTP (e.g. in the case of data from a web server running at the host) or IMAP (e.g. in the case of data from an email server running at the host).

The network interface device 102 includes a message engine 107 that is configured to process messages in accordance with an upper layer protocol. The message engine could receive data from the host computing device (for example, from application 106) or from an entity accessible over network 103 (e.g. a remote server) and process the received data in accordance with the upper layer protocol. The message engine is preferably controlled by means of a driver 108 at the operating system of the host device (which could be the driver of the NIC). The message engine could alternatively be supported at another peripheral device of system 100.

The first aspect of the present invention recognizes that, particularly in latency critical scenarios, it is often the case that upper layer messages are aligned with transport stream segment boundaries (e.g. TCP segment boundaries) such that each stream segment contains one or more complete upper layer messages. The first aspect of the present invention builds on this insight and provides a low latency messaging architecture that does not require a network interface device having a hardware offload engine configured to perform full transport stream protocol processing. This is achieved by arranging that the network interface device extracts payload data from the data packets of a transport stream and passes the payload data to message engine 107 for upper layer protocol processing. On the optimistic assumption that much of the time the upper layer messages will be aligned with the transport stream boundaries, the message engine will be able to process a useful proportion of the upper layer messages without transport stream processing having been completed.

Typically a predetermined set of transport streams will carry upper layer protocol messages that are to be handled if possible at message engine 107. In the following example, these transport streams will be referred to as TCP streams, but could in fact be transport streams of any protocol. Thus, network interface device 102 is configured to direct to the message engine the payload data of any data packets received at the NIC that belong to the predetermined set of TCP streams. This could be achieved by an entity of the NIC that is not the message engine extracting the payload data from the appropriate TCP packets and passing it to the message engine; alternatively, the message engine could have access to TCP data packets of the predetermined streams that are received at the NIC and itself extract the payload data for attempted upper layer processing.

As mentioned above, protocol layers below the transport layer (e.g. Ethernet and IP layers) could be processed at the NIC, but the present invention does not require that this is the case and preferably no substantial processing of protocol layers other than the upper layer protocol layer is performed at the NIC. Typically some protocol processing of lower layers will be performed at the NIC, such as checking or calculating checksums, or performing encryption/decryption such as IPSEC. The present invention requires that some part of the NIC can identify the payload data of a received data packet that represents messages according to the upper layer protocol. This can be readily achieved by identifying in a data packet those headers/footers/special characters/or other identifying data that delineates payload data of the upper layer protocol. In terms of lower layer protocol processing, the logic required at the NIC can therefore be limited to parsing a data packet to find the start of the payload data. Note that there may be one or more protocol layers between the TCP protocol and the upper layer protocol: in this case any headers, footers or other data relating to the intervening protocols does not form part of the payload data of the upper layer protocol that is passed to the message engine.

On receiving payload data of the upper layer protocol from a TCP data packet, the message engine attempts to process the payload data as if it were one or more upper layer messages. If the upper layer message starts at the start of the payload data and there are an integer number of messages in the payload data, one can expect that the upper layer processing of the payload data will be successful. If the processing is successful, the message engine causes to be passed to the host device one or both of (a) a signal indicating that upper layer processing was successfully performed, and (b) the data formed by the upper layer processing. If the processing is unsuccessful, the message engine signals this result to the host device so that the host knows that the message engine was not able to successfully perform upper layer processing for that data packet.

Upper layer processing could be unsuccessful because the payload data does not contain any messages, because the payload data is corrupted, because messages are received out of order by the message engine, or because only parts of a message are received (e.g. a message could span two or more data packets), because a checksum has failed (at any protocol layer), or because any lower layer frame or segment is fragmented (e.g. the stream could be a TCP/IP stream and the IP frame could be fragmented). This list is not exhaustive and upper layer processing could be unsuccessful for other reasons that would mean upper layer processing should not be performed or the results of any processing should be abandoned. Note that upper layer processing is unsuccessful for a given data packet if upper layer processing starts but does not complete due to an error, if upper layer processing completes but the results are to be discarded (e.g. due to a message being processed out of order), or if upper layer processing is prevented from starting (e.g. due to the respective IP frame being fragmented).

An illustration of a data packet 200 carrying two messages 206 and 207 in the payload data 204 of the packet is shown in FIG. 2. The data packet comprises a set of headers 202 that includes a TCP header 203 adjacent to the start of the payload data. Hence, by parsing the data packet and locating the end of TCP header 203, NIC 102 can identify the payload data of the packet comprising the upper layer messages 206 and 207. The packet may additionally include one or more footers 205 that the NIC is further configured to exclude from the payload data it passes to message engine 107.

Additionally, irrespective of whether or not the message engine successfully processes upper layer messages carried in the payload data, the network interface device is configured to pass at least the TCP headers of each data packet to the host so as to allow the host to perform TCP processing of data packets at stack 105. Assuming any data packet integrity checks are performed in hardware at the NIC (e.g. checksum calculations), only the TCP headers are required for stack 105 to complete TCP processing. Optionally, the NIC passes to the host the entire unprocessed TCP segment, including the payload data. The NIC is also configured to pass the headers of any other protocol layers of the data packet for which protocol processing is performed at the host.

In this manner, the message engine is configured to perform upper layer processing of received data packets whenever it is able to. Since the message engine can perform upper layer processing faster than the host stack, a low latency receive data path is provided requiring that the NIC also implement transport stream protocol processing. Any data packets that were not successfully processed by the message engine are at least partly handled at the appropriate upper layer protocol processing entity of the host (e.g. at protocol stack 105): upper layer protocol processing could either be performed in full at the host, or the host could extract the upper layer protocol messages that have not been validly processed at the NIC and pass those back to the message engine for processing once, for example, reassembly of lower layer protocol segments has been completed.

In the case in which the message engine successfully processes messages carried in a data packet, the host is preferably configured to check whether the processing performed by the message engine is valid in light of the TCP processing performed at the host. For example, the message engine assumes that it receives upper layer messages in order, which might not be true if data packets are received out of sequence. Since a given message might be processed in dependence on upper layer protocol state that is modified from one message to the next, processing messages in the wrong order could lead to messages being processed in dependence on incorrect state. If the host determines any processing performed by the message engine to be invalid, the respective data generated by the message engine is discarded and either the host performs the upper layer processing, or the host causes the message engine to re-perform the upper layer processing in respect of the affected messages. These checks are preferably performed at the network protocol stack 105 at the host, and are preferably performed in response to TCP processing being completed by the stack.

Preferably the message engine and/or network interface device are configured to pass the results of upper layer processing performed at the message engine in packet order along with at least part of the data packet (i.e. at least the transport stream headers). Thus, the network interface device and message engine are preferably configured to transmit at least part of the data packet (and optionally all) to the host along with the results of any upper layer processing performed at the message engine on its payload data (e.g. success/failure messages, and the output of the upper layer processing); subsequently the next data packet and the results of its upper layer processing are sent to the host. By interleaving data packets and the results of upper layer processing, the host can keep track of which upper layer processing results relate to which data packet. Alternatively, each data packet and the results of its upper layer processing are labelled by an identifier (for example, an IP address and port number, or a packet sequence number) so as to allow the host to associate the two; data packets and the results of upper layer processing can then be passed asynchronously to the host.

Data and messages may be passed between the network interface device/message engine and the host stack 105 by any suitable means, including by means of buffers, event queues, and out of band data.

In some implementations of the first aspect of present invention, the data generated by the upper layer processing of messages received over TCP streams will be passed to entities of the host device 101, such as an application 106, for use by those entities. In other implementations of the first aspect of the present invention, the results of upper layer protocol processing performed at the message engine could be used at the NIC or another hardware device.

A preferred embodiment of the first aspect of the present invention will now be described in which a limited set of state of the TCP protocol is maintained by the message engine so as to allow the engine to cope with messages that are not aligned to stream boundaries whilst still not requiring that the NIC supports a TCP protocol stack.

Each TCP data packet (or other transport stream packet) includes an identifier of the stream (e.g. IP address and port number) and a sequence number that identifies the position of the packet in the stream sequence. This information is held in the headers of a data packet and can be readily read from each data packet using simple logic. The stream identification information (e.g. IP address and port number) is used by the NIC to identify which data packets are to have their payload data processed by the message engine in accordance with the invention. In accordance with the preferred embodiment, the message engine 107 is further configured to maintain a sequence number for each stream that is to be handled by the message engine. This sequence number could be the sequence number of the last packet that was received over the respective stream, or it could be the sequence number of the next packet that is expected to be received for that stream. The sequence number of each data packet received over one of the predetermined streams could be written to the message engine by the NIC, preferably as the payload data is passed to the message engine. Alternatively, the message engine could be configured to snoop the sequence number of each received data packet whilst it is held at the NIC.

The preferred embodiment of the first aspect of the present invention will now be described with reference to FIG. 3, which illustrates a pair of data packets 300 and 310 that are received consecutively at the NIC over a particular TCP stream. The two packets carry three upper layer messages 306, 307 and 308, with message 307 spanning the two packets. The network interface device 102 is configured to parse each data packet so as to identify their TCP headers 303 and 313 in their packet headers 302 and 312, and hence identify the location of the payload data 304 and 314 in the data packets. The packets may also include footers 305 and 315.

On passing payload data 304 of packet 300 to the message engine, the network interface device also passes the TCP sequence number of that data packet, which the message engine maintains for the respective stream. The message engine identifies the first message 306 in the payload data and processes that message in accordance with the upper layer protocol, as has been described with respect to FIG. 2. The message engine then identifies a partial message 307: depending on the nature of the upper layer processing it may be possible to start processing of the partial message and then retain meta-data describing the processing state of the message engine until the next part of the message is received; alternatively, the message engine may retain the partial message and defer processing until the rest of the message has been received. Message parts and/or meta-data describing the processing state of the message engine could be stored at the message engine/NIC, or at a memory at the host. On receiving payload data 314 and the sequence number of the respective data packet 310, the message engine checks that the sequence number of the respective data packet 310 indicates that it is the next in the sequence. If the packets have been received in order, the message engine updates the TCP sequence number it maintains for that stream and proceeds to identify the remaining part of message 307 in the payload data 314. Once message 307 has been reformed using the stored first part of the message, the message engine completes upper layer protocol processing of message 307 (if necessary using the stored upper layer protocol state) and subsequently performs upper layer processing of message 308.

Other aspects of the preferred embodiment operate in the manner described above. For example, if the message engine encounters an error, or the sequence numbers indicate that TCP segments have been received out of order, the message engine signals to the host that upper layer processing of the respective data packet payloads was not successful. The upper layer processing is then either performed at stack 105 following TCP processing of the respective data packet, or the stack passes messages extracted from reassembled TCP segments back to the message engine for processing.

It can be advantageous to allow the NIC to pass decoded upper layer messages to the entity that requires those messages before the host stack has completed protocol processing of the TCP and lower layers. For example, the message engine could be configured to write the decoded messages into an application buffer or the memory of a hardware offload engine at the system for use at the respective application/hardware engine. This allows the entity to which the messages are directed to use the decoded messages as soon as the messages have been decoded—i.e. at the latency of the message engine hardware, and not at the latency of the host software stack.

A second aspect of the present invention will now be described with reference to FIG. 1. The second aspect of the present invention relates to the transmission of upper layer messages generated at the message engine over transport streams by means of the network interface device. Message engine 107 is configured to generate upper layer messages from data that could be generated at the message engine, at another peripheral device of the data processing system 100, or at a software entity of the host device (e.g. application 106). Network interface device 102 is configured to transmit the upper layer messages in data packets over one of the predetermined set of transport streams but the NIC does not include a full TCP stack. Instead, when an upper layer message is to be transmitted, the network protocol stack 105 performs TCP and lower layer transmit processing so as to generate the information required to form the TCP and lower layer headers of the data packet for carrying the upper layer message. This information is passed to the network interface device (preferably the message engine) so as to enable the data packet to be generated for carrying an upper layer message. Typically the necessary TCP information would include one or more of: TCP receive window size, ACK number, segment sequence number, source and destination port numbers, maximum segment size, and any options set for the particular TCP stream. Analogous parameters exist for other transport stream protocols. The information necessary to form lower layer protocol headers might include (for a TCP/IP/Ethernet data packet) Ethernet MAC address, VLAN-ids, IP addresses and IP IDs. The host stack could provide the information by passing a template data packet to the NIC that includes the necessary headers and into which the upper layer message payload can be inserted.

Preferably, the NIC/message engine is configured to request in advance the information required from the host stack 105 to form one or more TCP data packets for carrying the messages so that the NIC can form each TCP data packet as soon as the message engine generates each upper layer message. This allows the upper layer messages to be transmitted at the lowest possible latency and ensures that messages are transmitted at a rate determined by the performance of the message engine.

On an upper layer message being ready for transmission, the network interface device forms the transport layer protocol headers in dependence on the information received from the host stack 105. If the upper layer message to be transmitted can be carried in a single stream segment, the NIC forms the stream segment, includes the message as the payload of the data packet, and transmits the data packet over the appropriate stream. Optionally, the network interface device further sends a copy of the transmitted packet—or just the payload data it has generated for the packet—to the host so as to allow the host stack to handle retransmission requests. Each transmitted data packet could be, for example, written to a buffer in host memory that is accessible to the host stack.

If an upper layer message that is to be transmitted by the NIC is too large to be carried by a single stream segment, the NIC generates multiple data packets sufficient to carry the upper layer message. This can be achieved using the TCP information transmitted to it without requiring the NIC to support a TCP stack. Typically the only parameter the NIC is required to modify in the TCP header from one packet to the next is the sequence number, which the NIC is configured to increment for each additional data packet that it transmits. The NIC is configured to send the updated sequence numbers to the host so as to allow the host stack to update its transmit sequence number for the respective TCP endpoint. This can be achieved by sending a copy of the transmitted data packet to the host and/or transmitting the sequence numbers or payload length of TCP segments transmitted. The NIC could be configured to explicitly signal to the host stack that it has generated additional data packets for the respective TCP stream and notify by means of the signalling the sequence numbers of the additional data packets sent/the total length of payload data transmitted.

So as to ensure that the NIC does not exceed the TCP receive window for a given TCP stream, the host stack preferably signals to the NIC how many packets, or the total length of payload data, that the NIC is permitted to transmit over that stream before making a further request to the host stack for permission (and the necessary information) to transmit data packets over the respective TCP stream.

If retransmission of a data packet formed at the NIC is required, the host stack (which handles the receive path and is therefore notified of the need for a retransmission) can cause retransmission of the appropriate data packet in one of three ways, depending on the information available to it:

The host stack could modify the copy it has of the originally-transmitted data packet that was passed to it by the NIC and transmit the modified copy in a conventional manner by means of the NIC.

The host stack could recreate the data packet from the original input data and TCP parameters provided to the message engine; the upper layer messages can be recreated by the host stack from the original input data and the TCP parameters can be used to form the TCP segment.

The host stack could pass to the message engine the original input data and TCP information so as to cause the message engine to recreate the upper layer messages and allow the NIC to form the TCP segment according to the principles of the present invention.

Note that the source data provided to the message engine for processing into upper layer messages could come from any source, including: from software at the host (e.g. application 106); from the NIC itself (e.g. an offload engine at the NIC, such as a financial trading FPGA); from another device (e.g. another peripheral device of the system). If the source data does not originate at the host (i.e. the data originates at the NIC or another device), the NIC is configured to write to the host at least the payload data (comprising the upper layer messages it has formed) of each data packet it transmits so as to allow the host to perform retransmission of stream segments. If the source data does originate at the host, this may not be required if the host stack has access to the original source data.

In this manner, the substantial logic and memory required to implement a TCP stack in hardware at a NIC is not required, and the host is not required to send the full TCP control block to the NIC. However, the ability of the NIC to form and transmit TCP data packets comprising upper layer messages generated at the message engine is provided without the NIC having to wait for the relatively high latency software stack to generate the required TCP data packets. The formation of TCP data packets is preferably performed at message engine 107, with the information required to generate TCP data packets being passed to the message engine by means of driver 108.

A software entity for managing the NIC could be used to configure which transport streams have their receive and/or transmit paths handled in accordance with the present invention—for example a software entity such as application 106 could indicate to the host stack that all transport streams established with a particular remote server are to be handled according to the low latency mechanisms described herein whenever possible. Alternatively, the host stack could be arranged to use the low latency mechanisms for all transport streams, or all those that meet a predetermined set of criteria.

A message engine 107 of the present invention could be provided as a field programmable gate array (FPGA) at the NIC 102, or at another peripheral device that is arranged to communicate with the NIC over a bus or other interconnect. Such another peripheral device can be considered to form part of NIC 102 shown in FIG. 1.

A first exemplary implementation of the present invention will now be described. Host computing device 101 supports a trading application 106 configured to interact with a remote electronic exchange accessible to the host over one or more TCP streams established over network 103 by means of network protocol stack 105 and network interface device 102. Once a trading strategy has been selected by means of application 106, the host stack configures the message engine of the NIC to place trades at the electronic exchange over the one or more TCP streams under the control of a set of trading algorithms configured to effect the selected trading strategy. The trading algorithms could be supported in software at the host (e.g. at the trading application) or in hardware at a trading engine (which could be at the NIC, and could be part of the message engine). The trades are transmitted to the exchange as messages processed by the message engine in accordance with the FIX (financial information exchange) upper layer protocol and sent over one of the TCP streams. The transmission of these messages is performed in accordance with the second aspect of the present invention.

The message engine would preferably be an FPGA that supports both FIX protocol processing and the trading algorithms configured to effect the selected trading strategy. The trading decisions taken by algorithms at the message engine could be informed by financial data received over the network (e.g. by FIX-Fast over UDP) and processed at dedicated hardware of the NIC (again, this could be at the message engine) or another peripheral device of the system.

Responses from the exchange to the trades placed by the message engine are received as FIX messages over TCP in accordance with the first embodiment of the present invention. Thus, messages from the exchange are decoded at the message engine and delivered to the host. The host stack performs TCP protocol processing, handles any exception cases (e.g. where the FIX processing at the message engine was unsuccessful) and releases the decoded FIX messages to the trading software and/or the trading algorithms so as to inform those entities whether the trades were successful.

Since the upper layer messages received from the remote electronic exchange are mediated over TCP streams configured for low latency, it is likely that messages from the exchange will be aligned with the TCP boundaries of the stream over which they are carried. A significant proportion of the FIX messages received at NIC 102 are therefore likely to be successfully processed in hardware at the message engine without a full TCP stack being provided at the NIC, and without trading decisions being required to wait on relatively slow software stack 105 to perform TCP and upper layer processing. Thus, by performing the trades in hardware at the message engine as close as possible to the physical network, latency is reduced over a software implementation of upper layer processing.

A second exemplary implementation of the present invention will now be described. In this implementation, the data processing system 100 is configured to act as a broker or risk-management agent that operates on a set of TCP streams established between clients and an electronic exchange. A broker or risk-management agent mediates the exchange of FIX messages between clients and an electronic exchange, and may modify the messages exchanged between the two. In accordance with the first aspect of the present invention, the message engine of the NIC of the system is configured to decode each FIX message carried in a TCP data packet and pass the decoded messages to the appropriate entity at the host—typically a host application that is performing the broker or risk-management agent functions. TCP processing is performed at the host stack as described herein.

Once the host application has processed the decoded message and performed any necessary modifications, it causes the message engine to encode the modified message and forward it over the appropriate TCP stream to the intended endpoint (e.g. the exchange or a client). This is performed in accordance with the second aspect of the present invention, with TCP processing being performed at the host, and the message engine being provided with sufficient information to form a TCP packet around the modified FIX message it generates. Typically, the host application would cause the message engine to perform the FIX protocol processing by making a call to the host stack in respect of the appropriate TCP stream. The host stack would recognise that that TCP stream is being handled by the message engine and the stack would therefore hand over FIX protocol processing to the message engine.

The applicant hereby discloses in isolation each individual feature described herein and any combination of two or more such features, to the extent that such features or combinations are capable of being carried out based on the present specification as a whole in the light of the common general knowledge of a person skilled in the art, irrespective of whether such features or combinations of features solve any problems disclosed herein, and without limitation to the scope of the claims. The applicant indicates that aspects of the present invention may consist of any such individual feature or combination of features. In view of the foregoing description it will be evident to a person skilled in the art that various modifications may be made within the scope of the invention.

Claims (40)

What is claimed is:
1. A data processing system comprising:
a host computing device supporting an operating system and a network protocol stack, the network protocol stack being operable to support one or more transport streams by performing transport stream protocol processing of data packets received over the streams; and
a network interface device arranged to couple the host computing device to a network and operable to receive data packets over a transport stream supported by the network protocol stack, said network interface device comprising a message engine configured to perform upper layer protocol processing,
wherein the network interface device is configured to, on receiving a data packet over one of a predetermined set of transport streams, pass the payload data of the data packet to the message engine and the message engine is configured to, in response to receiving the payload data, identify and process any upper layer messages in the payload data in accordance with the upper layer protocol,
wherein said message engine processes said upper layer messages before the transport protocol processing have been completed, and is further configured to pass at least the headers of the data packet to the network protocol stack so as to cause the network protocol stack to perform transport protocol processing of the received data packet.
2. A data processing system as claimed in claim 1, wherein the network protocol stack is further configured to, in response to receiving at least the headers of the data packet, perform lower layer protocol processing of the received data packet so as to complete protocol processing of network protocols below the transport layer.
3. A data processing system as claimed in claim 1, wherein the network interface device passes the payload data to the message engine by making available at least the payload data to the message engine, and optionally the received data packet including its headers and the payload data.
4. A data processing system as claimed in claim 1, wherein the network interface device or message engine parses the received data packet so as to identify its payload data.
5. A data processing system as claimed in claim 1, wherein the payload data does not comprise any data relating to intervening protocol layers between the upper layer protocol and the transport stream protocol.
6. A data processing system as claimed in claim 1, wherein the payload data only comprises data relating to the upper layer protocol.
7. A data processing system as claimed in claim 1, wherein the message engine is configured to, if the upper layer processing is successful, pass data generated by the upper layer processing to the upper layer endpoint to which the upper layer messages are directed.
8. A data processing system as claimed in claim 7, wherein the message engine passes data generated by the upper layer processing to the network protocol stack by writing the data generated by the upper layer processing to a buffer at the host computing device that is accessible to the network protocol stack.
9. A data processing system as claimed in claim 1, wherein the message engine is configured to, if the upper layer processing is successful, pass data generated by the upper layer processing to the network protocol stack and the network protocol stack is configured to release the generated data to the upper layer endpoint to which the upper layer messages are directed only if the transport stream processing of the data packet is successful and the data packet is determined to have been received in-sequence.
10. A data processing system as claimed in claim 1, wherein the network protocol stack is configured to, if the network protocol stack determines that the data packet was received out-of-sequence, discard the data generated by the message engine and cause the upper layer protocol processing of the data packet to be performed in software at the host computing device.
11. A data processing system as claimed in claim 1, wherein the message engine is configured to, if the upper layer processing is unsuccessful, signal the network protocol stack to indicate that upper layer processing was unsuccessful, and the network protocol stack is configured to, in response to the signal, cause the upper layer protocol processing of the data packet to be performed in software at the host computing device.
12. A data processing system as claimed in claim 11, wherein the upper layer protocol processing is performed in software at the network protocol stack or at an application supported by the operating system of the host computing device.
13. A data processing system as claimed in claim 1, wherein the message engine is further configured to maintain a packet sequence number for each of the predetermined set of transport streams and to, if the payload data comprises a partial upper layer message, perform the following steps:
process the partial upper layer message in accordance with the upper layer protocol;
store metadata describing the processing state of the message engine; and
on the network interface device receiving a subsequent data packet over the one of the predetermined set of transport streams, check that the packet sequence number of the subsequent data packet indicates that it is the next data packet in the sequence of data packets transmitted over that transport stream, and if so:
load the metadata describing the processing state of the message engine and continue upper layer protocol processing of the respective upper layer message; and
otherwise, abandon the upper layer protocol processing of the respective upper layer message.
14. A data processing system as claimed in claim 13, wherein the message engine is further configured to, on upper layer protocol processing of the respective upper layer message being completed, pass data generated by the upper layer processing to the endpoint to which the upper layer messages are directed.
15. A data processing system as claimed in claim 13, wherein the message engine is further configured to, if upper layer protocol processing of the respective upper layer message is abandoned, signal to the network protocol stack so as to cause the upper layer protocol processing of the respective data packets to be performed in software at the host computing device.
16. A data processing system as claimed in claim 1, wherein the upper layer endpoint is an application supported by the operating system of the host computing device, the message engine itself, or another hardware engine of the data processing system.
17. A data processing system as claimed in claim 16, wherein the another hardware engine is a financial trading engine provided at a peripheral device of the host computing system and the upper layer protocol is FIX.
18. A data processing system as claimed in claim 1, wherein the lower layer protocols include one or more of IP, IPSec, IEEE 802.3, and IEEE 802.11.
19. A data processing system as claimed in claim 1, wherein the network interface device is not operable to perform transport stream protocol processing to completion.
20. A data processing system as claimed in claim 1, wherein the network protocol stack is a user level network protocol stack.
21. A data processing system as claimed in claim 1, wherein the transport stream protocol is TCP.
22. A data processing system comprising:
a host computing device supporting an operating system and a network protocol stack, the network protocol stack being operable to support one or more transport streams by performing transport stream protocol processing of data packets received over the streams; and
a network interface device arranged to couple the host computing device to a network and operable to transmit data packets over a transport stream supported by the network protocol stack, said network interface device comprising a message engine configured to perform upper layer protocol processing,
wherein the network protocol stack is configured to, on upper layer data being for transmission over one of a predetermined set of transport streams, form transport stream information sufficient for the message engine to form a data packet without performing transport stream protocol processing and cause the message engine to process the upper layer data in accordance with the upper layer protocol so as to form one or more upper layer messages; and the message engine being configured to, in dependence on the transport stream information, encapsulate the one or more upper layer messages in one or more data packets and cause the network interface device to transmit the data packet over the transport stream.
23. A data processing system as claimed in claim 22, wherein the transport stream protocol is TCP and the transport stream information includes: segment sequence number, TCP receive window size, ACK number, source port number, destination port number, maximum segment size, and the TCP options of the TCP stream.
24. A data processing system as claimed in claim 22, wherein the upper layer data for transmission is generated at the message engine, at an application supported at the operating system of the host computing device, or at another hardware engine of the data processing system.
25. A data processing system as claimed in claim 24, wherein the another hardware engine is a financial trading engine provided at a peripheral device of the host computing system and the upper layer protocol is FIX.
26. A data processing system as claimed in claim 22, wherein the transport stream information comprises information sufficient for the message engine to form the data packet without performing lower layer protocol processing.
27. A data processing system as claimed in claim 26, wherein the transport stream information includes one or more of an Ethernet MAC address, a VLAN-id, an IP address and an IP ID.
28. A data processing system as claimed in claim 22, wherein the transport stream information comprises a data packet template.
29. A data processing system as claimed in claim 22, wherein the message engine is configured to request transport stream information from the network protocol stack on the message engine receiving upper layer data for transmission.
30. A data processing system as claimed in claim 22, wherein the transport stream information includes a packet sequence number and the message engine is configured to, if the one or more upper layer messages require more than one data packet, for each additional data packet increment the packet sequence number.
31. A data processing system as claimed in claim 22, wherein the message engine is configured to make available to the network protocol stack each data packet it causes to be transmitted.
32. A data processing system as claimed in claim 31, wherein the network protocol stack is configured to, if a data packet is to be retransmitted over one of the predetermined set of transport streams, effect retransmission of that data packet by modifying a copy of the data packet made available to it by the message engine and causing the modified copy to be transmitted by the network interface device.
33. A data processing system as claimed in claim 22, wherein the message engine is configured to signal to the network protocol stack the packet sequence numbers of any additional data packets it causes to be transmitted.
34. A data processing system as claimed in claim 22, wherein the network protocol stack is further configured to signal to the message engine the amount of payload data the message engine is permitted to cause to be transmitted by the network interface device.
35. A data processing system as claimed in claim 34, wherein the amount of payload data is expressed as a number of data packets or a length of payload data.
36. A data processing system as claimed in claim 22, wherein the network protocol stack has access to the upper layer data for transmission and is configured to, if a data packet is to be retransmitted over one of the predetermined set of transport streams, effect retransmission of that data packet by:
passing the respective upper layer data for transmission to the message engine along with transport stream information sufficient to generate the data packet for retransmission; or
passing the respective upper layer data for transmission to a software entity of the host computing device so as to cause that entity to form the respective one or more upper layer messages and forming the data packet for retransmission at the network protocol stack.
37. A data processing system as claimed in claim 22, wherein the lower layer protocols include one or more of IP, IPSec, IEEE 802.3, and IEEE 802.11.
38. A data processing system as claimed in claim 22, wherein the network interface device is not operable to perform transport stream protocol processing to completion.
39. A data processing system as claimed in claim 22, wherein the network protocol stack is a user level network protocol stack.
40. A data processing system as claimed in claim 22, wherein the transport stream protocol is TCP.
US13/624,788 2011-09-22 2012-09-21 Message acceleration Active 2033-04-22 US9003053B2 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US201161538055P true 2011-09-22 2011-09-22
US13/624,788 US9003053B2 (en) 2011-09-22 2012-09-21 Message acceleration

Applications Claiming Priority (5)

Application Number Priority Date Filing Date Title
US13/624,788 US9003053B2 (en) 2011-09-22 2012-09-21 Message acceleration
US13/754,792 US20130145035A1 (en) 2010-12-09 2013-01-30 Tcp processing for devices
US14/231,510 US9674318B2 (en) 2010-12-09 2014-03-31 TCP processing for devices
US15/287,666 US20170026497A1 (en) 2010-12-09 2016-10-06 Tcp processing for devices
US15/481,350 US20170214775A1 (en) 2010-12-09 2017-04-06 Tcp processing for devices

Related Parent Applications (2)

Application Number Title Priority Date Filing Date
US13/283,420 Continuation-In-Part US9258390B2 (en) 2011-07-29 2011-10-27 Reducing network latency
US13/671,434 Continuation-In-Part US9600429B2 (en) 2010-12-09 2012-11-07 Encapsulated accelerator

Related Child Applications (2)

Application Number Title Priority Date Filing Date
US13/283,420 Continuation-In-Part US9258390B2 (en) 2011-07-29 2011-10-27 Reducing network latency
US13/754,792 Continuation-In-Part US20130145035A1 (en) 2010-12-09 2013-01-30 Tcp processing for devices

Publications (2)

Publication Number Publication Date
US20130080651A1 US20130080651A1 (en) 2013-03-28
US9003053B2 true US9003053B2 (en) 2015-04-07

Family

ID=47189707

Family Applications (1)

Application Number Title Priority Date Filing Date
US13/624,788 Active 2033-04-22 US9003053B2 (en) 2011-09-22 2012-09-21 Message acceleration

Country Status (2)

Country Link
US (1) US9003053B2 (en)
EP (1) EP2574000A3 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9880964B2 (en) 2010-12-09 2018-01-30 Solarflare Communications, Inc. Encapsulated accelerator

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9444914B2 (en) 2013-09-16 2016-09-13 Annapurna Labs Ltd. Configurable parser and a method for parsing information units
RU2013158632A (en) * 2013-12-30 2015-07-10 Общество С Ограниченной Ответственностью "Яндекс" A method of data providing apparatus (embodiments), a method of establishing a communication session between the device and the web server and the server for providing data
US9444827B2 (en) * 2014-02-15 2016-09-13 Micron Technology, Inc. Multi-function, modular system for network security, secure communication, and malware protection
US20160330301A1 (en) 2015-05-07 2016-11-10 Mellanox Technologies Ltd. Efficient transport flow processing on an accelerator
US10152441B2 (en) 2015-05-18 2018-12-11 Mellanox Technologies, Ltd. Host bus access by add-on devices via a network interface controller
JP6470166B2 (en) * 2015-12-18 2019-02-13 日本電信電話株式会社 Order control unit, the order control method and program
US20180124216A1 (en) * 2016-11-02 2018-05-03 Solarflare Communications, Inc. Network interface device and host processing device field

Citations (168)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5272599A (en) 1993-03-19 1993-12-21 Compaq Computer Corporation Microprocessor heat dissipation apparatus for a printed circuit board
US5325532A (en) 1992-09-25 1994-06-28 Compaq Computer Corporation Automatic development of operating system boot image
EP0620521A2 (en) 1993-04-16 1994-10-19 Hewlett-Packard Company Method for dynamically building dialog boxes
US5946189A (en) 1997-04-23 1999-08-31 Compaq Computer Corporation Pivotable support and heat sink apparatus removably connectable without tools to a computer processor
WO2000010095A1 (en) 1998-08-14 2000-02-24 Intel Corporation Striping packets of data across multiple virtual channels
US6098112A (en) 1995-10-19 2000-08-01 Hewlett-Packard Company Streams function registering
US6160554A (en) 1998-03-19 2000-12-12 Hewlett Packard Company Computer file content preview window
WO2001048972A1 (en) 1999-12-28 2001-07-05 Mellanox Technologies Ltd. Adaptive sampling
US6304945B1 (en) 1999-05-13 2001-10-16 Compaq Computer Corporation Method and apparatus for maintaining cache coherency in a computer system having multiple processor buses
US6349035B1 (en) 2000-09-29 2002-02-19 Compaq Information Technologies Group, L.P. Method and apparatus for tooless mating of liquid cooled cold plate with tapered interposer heat sink
WO2002035838A1 (en) 2000-10-28 2002-05-02 Voltaire Limited Modular set top box assembly
US20020059052A1 (en) 2000-09-12 2002-05-16 Mellanox Technologies Ltd. Co-simulation of network components
US6427173B1 (en) 1997-10-14 2002-07-30 Alacritech, Inc. Intelligent network interfaced device and system for accelerated communication
US20020112139A1 (en) 2001-02-13 2002-08-15 Krause Michael R. Highly available, monotonic increasing sequence number generation
US6438130B1 (en) 2001-06-28 2002-08-20 Mellanox Technologies Ltd. Forwarding database cache
US20020129293A1 (en) 2001-03-07 2002-09-12 Hutton John F. Scan based multiple ring oscillator structure for on-chip speed measurement
US20020140985A1 (en) 2001-04-02 2002-10-03 Hudson Kevin R. Color calibration for clustered printing
US20020156784A1 (en) 1999-04-08 2002-10-24 Hanes David H. Method for transferring and indexing data from old media to new media
US6502203B2 (en) 1999-04-16 2002-12-31 Compaq Information Technologies Group, L.P. Method and apparatus for cluster system operation
US20030007165A1 (en) 2001-07-05 2003-01-09 Hudson Kevin R. Direction-dependent color conversion in bidirectional printing
US20030033588A1 (en) 2001-01-29 2003-02-13 John Alexander System, method and article of manufacture for using a library map to create and maintain IP cores effectively
US6530007B2 (en) 1998-07-13 2003-03-04 Compaq Information Technologies Group, L.P. Method and apparatus for supporting heterogeneous memory in computer systems
US20030058459A1 (en) 2001-09-26 2003-03-27 Yifeng Wu Generalized color calibration architecture and method
US20030065856A1 (en) 2001-10-03 2003-04-03 Mellanox Technologies Ltd. Network adapter with multiple event queues
US20030063299A1 (en) 2001-10-02 2003-04-03 Cowan Philip B. Color calibration method and apparatus
US20030081060A1 (en) 2001-10-29 2003-05-01 Huanzhao Zeng Method and system for management of color through conversion between color spaces
US20030086300A1 (en) 2001-04-06 2003-05-08 Gareth Noyes FPGA coprocessing system
US20030172330A1 (en) 2002-03-11 2003-09-11 Barron Dwight L. Data redundancy in a hot pluggable, large symmetric multi-processor system
US20030191786A1 (en) 1999-02-01 2003-10-09 Matson Mark D. Computer method and apparatus for division and square root operations using signed digit
US20030202043A1 (en) 2002-04-24 2003-10-30 Huanzhao Zeng Determination of control points for construction of first color space-to-second color space look-up table
US20030214677A1 (en) 1998-03-26 2003-11-20 Ranjit Bhaskar Tone dependent variable halftoning wih adjustable algorithm selection
US6667918B2 (en) 2002-05-01 2003-12-23 Mellanox Technologies Ltd. Self-repair of embedded memory arrays
US6718392B1 (en) 2000-10-24 2004-04-06 Hewlett-Packard Development Company, L.P. Queue pair partitioning in distributed computer system
US20040071250A1 (en) 2000-06-20 2004-04-15 Bunton William P. High-speed interconnection adapter having automated lane de-skew
US6728743B2 (en) 2001-01-04 2004-04-27 Mellanox Technologies Ltd. Modulo remainder generator
US6735642B2 (en) 2000-06-02 2004-05-11 Mellanox Technologies Ltd. DMA doorbell
US20040141642A1 (en) 2003-01-22 2004-07-22 Huanzhao Zeng Color space conversion using interpolation
US6768996B1 (en) 1998-10-08 2004-07-27 Hewlett-Packard Development Company, L.P. System and method for retrieving an abstracted portion of a file without regard to the operating system of the current host computer
US20040156393A1 (en) 2003-02-12 2004-08-12 Silverback Systems, Inc. Architecture and API for of transport and upper layer protocol processing acceleration
US20040190557A1 (en) 2003-03-27 2004-09-30 Barron Dwight L. Signaling packet
US20040193825A1 (en) 2003-03-27 2004-09-30 Garcia David J. Binding a memory window to a queue pair
US20040190533A1 (en) 2003-03-27 2004-09-30 Prashant Modi Method and apparatus for performing connection management with multiple stacks
US20040193734A1 (en) 2003-03-27 2004-09-30 Barron Dwight L. Atomic operations
US20040190538A1 (en) 2000-09-08 2004-09-30 Hewlett-Packard Development Company, L.P. Speculative transmit for system area network latency reduction
US20040210754A1 (en) 2003-04-16 2004-10-21 Barron Dwight L. Shared security transform device, system and methods
US20040210670A1 (en) * 1999-03-05 2004-10-21 Nikolaos Anerousis System, method and apparatus for network service load and reliability management
US20040240435A1 (en) 1997-10-14 2004-12-02 Alacritech, Inc. Obtaining a destination address so that a network interface device can write network data without headers directly into host memory
US20040249881A1 (en) 2003-06-05 2004-12-09 Jha Ashutosh K. Transmitting commands and information between a TCP/IP stack and an offload unit
US20040252685A1 (en) 2003-06-13 2004-12-16 Mellanox Technologies Ltd. Channel adapter with integrated switch
US20050008223A1 (en) 2003-07-09 2005-01-13 Huanzhao Zeng Representing extended color gamut information
US20050018221A1 (en) 2003-07-21 2005-01-27 Huanzhao Zeng Conforming output intensities of pens
US20050038941A1 (en) 2003-08-14 2005-02-17 Mallikarjun Chadalapaka Method and apparatus for accessing a memory
US20050039171A1 (en) 2003-08-12 2005-02-17 Avakian Arra E. Using interceptors and out-of-band data to monitor the performance of Java 2 enterprise edition (J2EE) applications
US20050039172A1 (en) 2003-08-12 2005-02-17 Jeffrey Rees Synthesizing application response measurement (ARM) instrumentation
US20050038918A1 (en) 2003-08-14 2005-02-17 Hilland Jeffrey R. Method and apparatus for implementing work request lists
US20050039187A1 (en) 2003-08-12 2005-02-17 Avakian Arra E. Instrumenting java code by modifying bytecodes
US20050066333A1 (en) 2003-09-18 2005-03-24 Krause Michael R. Method and apparatus for providing notification
US6904534B2 (en) 2001-09-29 2005-06-07 Hewlett-Packard Development Company, L.P. Progressive CPU sleep state duty cycle to limit peak power of multiple computers on shared power distribution unit
US20050172181A1 (en) 2004-01-16 2005-08-04 Mellanox Technologies Ltd. System and method for production testing of high speed communications receivers
US20050219278A1 (en) 2004-03-30 2005-10-06 Hudson Kevin R Image formation with a flexible number of passes
US20050219314A1 (en) 2004-03-30 2005-10-06 David Donovan Formation of images
US20050231751A1 (en) 2004-04-15 2005-10-20 Yifeng Wu Image processing system and method
US6978331B1 (en) 1999-09-08 2005-12-20 Mellanox Technologies Ltd. Synchronization of interrupts with data packets
US20060026443A1 (en) 2004-07-29 2006-02-02 Mcmahan Larry N Communication among partitioned devices
US20060045098A1 (en) 2004-08-31 2006-03-02 Krause Michael R System for port mapping in a network
US20060126619A1 (en) 2004-12-14 2006-06-15 Teisberg Robert R Aggregation over multiple processing nodes of network resources each providing offloaded connections between applications over a network
US20060165074A1 (en) 2004-12-14 2006-07-27 Prashant Modi Aggregation of network resources providing offloaded connections between applications over a network
US7099275B2 (en) 2001-09-21 2006-08-29 Slt Logic Llc Programmable multi-service queue scheduler
US20060193318A1 (en) 2005-02-28 2006-08-31 Sriram Narasimhan Method and apparatus for processing inbound and outbound quanta of data
US7103626B1 (en) 2000-05-24 2006-09-05 Hewlett-Packard Development, L.P. Partitioning in distributed computer system
US20060228637A1 (en) 2005-04-11 2006-10-12 Jackson Bruce J Color filter
US20060248191A1 (en) 2005-04-27 2006-11-02 Hudson Charles L Aggregation of hybrid network resources operable to support both offloaded and non-offloaded connections
US7136397B2 (en) 2001-08-20 2006-11-14 Slt Logic Llc Network architecture and system for delivering bi-directional xDSL based services
US7143412B2 (en) 2002-07-25 2006-11-28 Hewlett-Packard Development Company, L.P. Method and apparatus for optimizing performance in a multi-processing system
US7149227B2 (en) 2002-05-31 2006-12-12 Mellanox Technologies Ltd. Round-robin arbiter with low jitter
US7151744B2 (en) 2001-09-21 2006-12-19 Slt Logic Llc Multi-service queuing method and apparatus that provides exhaustive arbitration, load balancing, and support for rapid port failover
US7216225B2 (en) 2000-05-24 2007-05-08 Voltaire Ltd. Filtered application-to-application communication
US7240350B1 (en) 2002-01-07 2007-07-03 Slt Logic, Llc System and method for providing communications to processes
US7245627B2 (en) 2002-04-23 2007-07-17 Mellanox Technologies Ltd. Sharing a network interface card among multiple hosts
US7254237B1 (en) 2001-01-12 2007-08-07 Slt Logic, Llc System and method for establishing a secure connection
US20070188351A1 (en) 2005-12-16 2007-08-16 Andrew Brown Hardware enablement using an interface
US20070220183A1 (en) 2002-07-23 2007-09-20 Mellanox Technologies Ltd. Receive Queue Descriptor Pool
US20070237327A1 (en) 2006-03-23 2007-10-11 Exegy Incorporated Method and System for High Throughput Blockwise Independent Encryption/Decryption
US7285996B2 (en) 2005-09-30 2007-10-23 Slt Logic, Llc Delay-locked loop
US20070260602A1 (en) 2006-05-02 2007-11-08 Exegy Incorporated Method and Apparatus for Approximate Pattern Matching
US20070277036A1 (en) 2003-05-23 2007-11-29 Washington University, A Corporation Of The State Of Missouri Intelligent data storage and processing using fpga devices
US7316017B1 (en) 2003-01-06 2008-01-01 Slt Logic, Llc System and method for allocatiing communications to processors and rescheduling processes in a multiprocessor system
US20080008205A1 (en) 2006-07-07 2008-01-10 Jung Byung Kwon DATA ACCELERATION APPARATUS FOR iSCSI AND iSCSI STORAGE SYSTEM USING THE SAME
US20080024586A1 (en) 2006-07-31 2008-01-31 Rodolfo Jodra Barron System and method for calibrating a beam array of a printer
US7346702B2 (en) 2000-08-24 2008-03-18 Voltaire Ltd. System and method for highly scalable high-speed content-based filtering and load balancing in interconnected fabrics
US20080109526A1 (en) 2006-11-06 2008-05-08 Viswanath Subramanian Rdma data to responder node coherency domain
US20080115217A1 (en) 2006-10-31 2008-05-15 Hewlett-Packard Development Company, L.P. Method and apparatus for protection of a computer system from malicious code attacks
US20080115216A1 (en) 2006-10-31 2008-05-15 Hewlett-Packard Development Company, L.P. Method and apparatus for removing homogeneity from execution environment of computing system
US20080126509A1 (en) 2006-11-06 2008-05-29 Viswanath Subramanian Rdma qp simplex switchless connection
US7386619B1 (en) 2003-01-06 2008-06-10 Slt Logic, Llc System and method for allocating communications to processors in a multiprocessor system
US20080140574A1 (en) 2002-11-06 2008-06-12 Boucher Laurence B Enabling an enhanced function of an electronic device
US20080135774A1 (en) 2006-12-08 2008-06-12 Asml Netherlands B.V. Scatterometer, a lithographic apparatus and a focus analysis method
US20080147828A1 (en) 2006-12-18 2008-06-19 Enstone Mark R Distributed Configuration Of Network Interface Cards
US20080148400A1 (en) 2006-10-31 2008-06-19 Hewlett-Packard Development Company, L.P. Method and apparatus for enforcement of software licence protection
US20080177890A1 (en) 1999-05-24 2008-07-24 Krause Michael R Reliable datagram via independent source destination resources
US20080244060A1 (en) 2007-03-30 2008-10-02 Cripe Daniel N Electronic device profile migration
WO2008127672A2 (en) 2007-04-11 2008-10-23 Slt Logic Llc Modular blade for providing scalable mechanical, electrical and environmental functionality in the enterprise using advanced tca boards
US20080304519A1 (en) 2007-06-06 2008-12-11 Hewlett-Packard Development Company, L.P. Method for ethernet power savings on link aggregated groups
US20090060197A1 (en) 2007-08-31 2009-03-05 Exegy Incorporated Method and Apparatus for Hardware-Accelerated Encryption/Decryption
US7502870B1 (en) * 2004-08-13 2009-03-10 Sun Microsystems, Inc. Method for receiving network communication and apparatus for performing the same
US7518164B2 (en) 2006-03-29 2009-04-14 Mellanox Technologies Ltd. Current-triggered low turn-on voltage SCR
US20090165003A1 (en) 2007-12-21 2009-06-25 Van Jacobson System and method for allocating communications to processors and rescheduling processes in a multiprocessor system
US20090182683A1 (en) 2008-01-11 2009-07-16 Exegy Incorporated Method and System for Low Latency Basket Calculation
US20090183057A1 (en) 2008-01-15 2009-07-16 Neterion , Inc. OFFLOADING iSCSI WITHOUT TOE
US7573967B2 (en) 2005-07-01 2009-08-11 Slt Logic Llc Input threshold adjustment in a synchronous data sampling circuit
US20090201926A1 (en) 2006-08-30 2009-08-13 Mellanox Technologies Ltd Fibre channel processing by a host channel adapter
US7580495B2 (en) 2005-06-30 2009-08-25 Slt Logic Llc Mixer-based phase control
US20090213856A1 (en) 2001-05-04 2009-08-27 Slt Logic Llc System and Method for Providing Transformation of Multi-Protocol Packets in a Data Stream
US20090268612A1 (en) 2008-04-29 2009-10-29 Google Inc. Method and apparatus for a network queuing engine and congestion management gateway
WO2009134219A1 (en) 2008-04-28 2009-11-05 Hewlett-Packard Development Company, L.P. Adjustable server-transmission rates over fixed-speed backplane connections within a multi-server enclosure
WO2009136933A1 (en) 2008-05-08 2009-11-12 Hewlett-Packard Development Company, L.P. A method for interfacing a fibre channel network with an ethernet based network
US20090287628A1 (en) 2008-05-15 2009-11-19 Exegy Incorporated Method and System for Accelerated Stream Processing
US7631106B2 (en) 2002-08-15 2009-12-08 Mellanox Technologies Ltd. Prefetching of receive queue descriptors
US20090302923A1 (en) 2006-05-08 2009-12-10 Mellanox Technologies Ltd. Terminated input buffer with offset cancellation circuit
US7653754B2 (en) 2004-01-05 2010-01-26 Mellanox Technologies Ltd. Method, system and protocol that enable unrestricted user-level access to a network interface adapter
WO2010020907A2 (en) 2008-08-21 2010-02-25 Voltaire Ltd. Device, system, and method of distributing messages
US7688853B2 (en) 2001-05-04 2010-03-30 Slt Logic, Llc System and method for hierarchical policing of flows and subflows of a data stream
US20100088437A1 (en) 2008-10-06 2010-04-08 Mellanox Technologies Ltd Infiniband adaptive congestion control adaptive marking rate
US7702629B2 (en) 2005-12-02 2010-04-20 Exegy Incorporated Method and device for high performance regular expression pattern matching
US7725556B1 (en) 2006-10-27 2010-05-25 Hewlett-Packard Development Company, L.P. Computer system with concurrent direct memory access
US20100138840A1 (en) 2005-08-23 2010-06-03 Mellanox Technologies Ltd. System and method for accelerating input/output access operation on a virtual machine
US20100169880A1 (en) 2008-12-25 2010-07-01 Voltaire Ltd. Virtual input-output connections for machine virtualization
US20100189206A1 (en) 2009-01-28 2010-07-29 Mellanox Technologies Ltd Precise Clock Synchronization
US20100188140A1 (en) 2006-04-07 2010-07-29 Mellanox Technologies Ltd. Accurate Global Reference Voltage Distribution System With Local Reference Voltages Referred To Local Ground And Locally Supplied Voltage
WO2010087826A1 (en) 2009-01-29 2010-08-05 Hewlett-Packard Development Company, L.P. Connecting ports of one or more electronic devices to different subsets of networks based on different operating modes
US7802071B2 (en) 2007-07-16 2010-09-21 Voltaire Ltd. Device, system, and method of publishing information to multiple subscribers
US7801027B2 (en) 2006-08-30 2010-09-21 Mellanox Technologies Ltd. Auto-negotiation by nodes on an infiniband fabric
US7813460B2 (en) 2005-09-30 2010-10-12 Slt Logic, Llc High-speed data sampler with input threshold adjustment
US20100265849A1 (en) 2009-04-21 2010-10-21 Voltaire Ltd Spanning tree root selection in a hierarchical network
US20100274876A1 (en) 2009-04-28 2010-10-28 Mellanox Technologies Ltd Network interface device with memory management capabilities
US7827442B2 (en) 2006-01-23 2010-11-02 Slt Logic Llc Shelf management controller with hardware/software implemented dual redundant configuration
US7835375B2 (en) 2001-05-04 2010-11-16 Slt Logic, Llc Method and apparatus for providing multi-protocol, multi-stage, real-time frame classification
US7835380B1 (en) 2004-10-19 2010-11-16 Broadcom Corporation Multi-port network interface device with shared processing resources
US7848322B2 (en) 2007-03-26 2010-12-07 Voltaire Ltd. Device, system and method of UDP communication
US7864787B2 (en) 2007-03-26 2011-01-04 Voltaire Ltd. Device, system and method of multicast communication
US20110004457A1 (en) 2006-08-07 2011-01-06 Voltaire Ltd. Service-oriented infrastructure management
US20110010557A1 (en) 2009-07-07 2011-01-13 Mellanox Technologies Ltd Control message signature for device control
US20110029847A1 (en) 2009-07-30 2011-02-03 Mellanox Technologies Ltd Processing of data integrity field
US20110040701A1 (en) 2006-06-19 2011-02-17 Exegy Incorporated Method and System for High Speed Options Pricing
US20110058571A1 (en) 2009-09-09 2011-03-10 Mellanox Technologies Ltd. Data switch with shared port buffers
US7921178B2 (en) 2008-12-04 2011-04-05 Voltaire Ltd. Device, system, and method of accessing storage
US20110083064A1 (en) 2009-10-04 2011-04-07 Mellanox Technologies Ltd Processing of block and transaction signatures
WO2011043769A1 (en) 2009-10-07 2011-04-14 Hewlett-Packard Development Company, L.P. Notification protocol based endpoint caching of host memory
US7930437B2 (en) 2001-04-11 2011-04-19 Mellanox Technologies Ltd. Network adapter with shared database for message context information
US7929539B2 (en) 2001-04-11 2011-04-19 Mellanox Technologies Ltd. Multiple queue pair access with a single doorbell
US20110096668A1 (en) 2009-10-26 2011-04-28 Mellanox Technologies Ltd. High-performance adaptive routing
US7934959B2 (en) 2009-10-06 2011-05-03 Mellanox Technologies Ltd. Adapter for pluggable module
WO2011053330A1 (en) 2009-11-02 2011-05-05 Hewlett-Packard Development Company, L.P. Multiprocessing computing with distributed embedded switching
WO2011053305A1 (en) 2009-10-30 2011-05-05 Hewlett-Packard Development Co Thermal bus bar for a blade enclosure
US20110113083A1 (en) 2009-11-11 2011-05-12 Voltaire Ltd Topology-Aware Fabric-Based Offloading of Collective Functions
US20110119673A1 (en) 2009-11-15 2011-05-19 Mellanox Technologies Ltd. Cross-channel network operation offloading for collective operations
US20110116512A1 (en) 2009-11-19 2011-05-19 Mellanox Technologies Ltd Dynamically-Connected Transport Service
US7954114B2 (en) 2006-01-26 2011-05-31 Exegy Incorporated Firmware socket module for FPGA-based pipeline processing
US7978606B2 (en) 2001-05-04 2011-07-12 Slt Logic, Llc System and method for policing multiple data flows and multi-protocol data flows
US20110173352A1 (en) 2010-01-13 2011-07-14 Mellanox Technologies Ltd Power Reduction on Idle Communication Lanes
US20110178918A1 (en) 2006-06-19 2011-07-21 Exegy Incorporated High Speed Processing of Financial Information Using FPGA Devices
US8156101B2 (en) 2006-11-13 2012-04-10 Exegy Incorporated Method and system for high performance integration, processing and searching of structured and unstructured data using coprocessors
US20120089496A1 (en) 2008-12-15 2012-04-12 Exegy Incorporated Method and Apparatus for High-Speed Processing of Financial Market Depth Data
US20120102245A1 (en) 2010-10-20 2012-04-26 Gole Abhijeet P Unified i/o adapter
EP2463782A2 (en) 2010-12-09 2012-06-13 Solarflare Communications Inc Encapsulated Accelerator
US20120246052A1 (en) 2010-12-09 2012-09-27 Exegy Incorporated Method and Apparatus for Managing Orders in Financial Markets
US8286193B2 (en) 2006-01-10 2012-10-09 Solarflare Communications, Inc. Processing received data
US8326816B2 (en) 2010-05-06 2012-12-04 Oracle International Corporation Fine grain synchronization for database replay

Patent Citations (199)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5325532A (en) 1992-09-25 1994-06-28 Compaq Computer Corporation Automatic development of operating system boot image
US5272599A (en) 1993-03-19 1993-12-21 Compaq Computer Corporation Microprocessor heat dissipation apparatus for a printed circuit board
EP0620521A2 (en) 1993-04-16 1994-10-19 Hewlett-Packard Company Method for dynamically building dialog boxes
US6098112A (en) 1995-10-19 2000-08-01 Hewlett-Packard Company Streams function registering
US5946189A (en) 1997-04-23 1999-08-31 Compaq Computer Corporation Pivotable support and heat sink apparatus removably connectable without tools to a computer processor
US6427173B1 (en) 1997-10-14 2002-07-30 Alacritech, Inc. Intelligent network interfaced device and system for accelerated communication
US20040240435A1 (en) 1997-10-14 2004-12-02 Alacritech, Inc. Obtaining a destination address so that a network interface device can write network data without headers directly into host memory
US6160554A (en) 1998-03-19 2000-12-12 Hewlett Packard Company Computer file content preview window
US20030214677A1 (en) 1998-03-26 2003-11-20 Ranjit Bhaskar Tone dependent variable halftoning wih adjustable algorithm selection
US6530007B2 (en) 1998-07-13 2003-03-04 Compaq Information Technologies Group, L.P. Method and apparatus for supporting heterogeneous memory in computer systems
WO2000010095A1 (en) 1998-08-14 2000-02-24 Intel Corporation Striping packets of data across multiple virtual channels
US6768996B1 (en) 1998-10-08 2004-07-27 Hewlett-Packard Development Company, L.P. System and method for retrieving an abstracted portion of a file without regard to the operating system of the current host computer
US20030191786A1 (en) 1999-02-01 2003-10-09 Matson Mark D. Computer method and apparatus for division and square root operations using signed digit
US20040210670A1 (en) * 1999-03-05 2004-10-21 Nikolaos Anerousis System, method and apparatus for network service load and reliability management
US20020156784A1 (en) 1999-04-08 2002-10-24 Hanes David H. Method for transferring and indexing data from old media to new media
US7509355B2 (en) 1999-04-08 2009-03-24 Hewlett-Packard Development Company, L.P. Method for transferring and indexing data from old media to new media
US6502203B2 (en) 1999-04-16 2002-12-31 Compaq Information Technologies Group, L.P. Method and apparatus for cluster system operation
US6304945B1 (en) 1999-05-13 2001-10-16 Compaq Computer Corporation Method and apparatus for maintaining cache coherency in a computer system having multiple processor buses
US7904576B2 (en) 1999-05-24 2011-03-08 Hewlett-Packard Development Company, L.P. Reliable datagram via independent source destination resources
US20080177890A1 (en) 1999-05-24 2008-07-24 Krause Michael R Reliable datagram via independent source destination resources
US6978331B1 (en) 1999-09-08 2005-12-20 Mellanox Technologies Ltd. Synchronization of interrupts with data packets
WO2001048972A1 (en) 1999-12-28 2001-07-05 Mellanox Technologies Ltd. Adaptive sampling
US7103626B1 (en) 2000-05-24 2006-09-05 Hewlett-Packard Development, L.P. Partitioning in distributed computer system
US7216225B2 (en) 2000-05-24 2007-05-08 Voltaire Ltd. Filtered application-to-application communication
US6735642B2 (en) 2000-06-02 2004-05-11 Mellanox Technologies Ltd. DMA doorbell
US20040071250A1 (en) 2000-06-20 2004-04-15 Bunton William P. High-speed interconnection adapter having automated lane de-skew
US7346702B2 (en) 2000-08-24 2008-03-18 Voltaire Ltd. System and method for highly scalable high-speed content-based filtering and load balancing in interconnected fabrics
US20040190538A1 (en) 2000-09-08 2004-09-30 Hewlett-Packard Development Company, L.P. Speculative transmit for system area network latency reduction
US20020059052A1 (en) 2000-09-12 2002-05-16 Mellanox Technologies Ltd. Co-simulation of network components
US6349035B1 (en) 2000-09-29 2002-02-19 Compaq Information Technologies Group, L.P. Method and apparatus for tooless mating of liquid cooled cold plate with tapered interposer heat sink
US6718392B1 (en) 2000-10-24 2004-04-06 Hewlett-Packard Development Company, L.P. Queue pair partitioning in distributed computer system
WO2002035838A1 (en) 2000-10-28 2002-05-02 Voltaire Limited Modular set top box assembly
US6728743B2 (en) 2001-01-04 2004-04-27 Mellanox Technologies Ltd. Modulo remainder generator
US7254237B1 (en) 2001-01-12 2007-08-07 Slt Logic, Llc System and method for establishing a secure connection
US20040015502A1 (en) 2001-01-29 2004-01-22 John Alexander Application program interface for programmable architecture cores
US20030033588A1 (en) 2001-01-29 2003-02-13 John Alexander System, method and article of manufacture for using a library map to create and maintain IP cores effectively
US6950961B2 (en) 2001-02-13 2005-09-27 Hewlett-Packard Development Company, L.P. Highly available, monotonic increasing sequence number generation
US20020112139A1 (en) 2001-02-13 2002-08-15 Krause Michael R. Highly available, monotonic increasing sequence number generation
US20020129293A1 (en) 2001-03-07 2002-09-12 Hutton John F. Scan based multiple ring oscillator structure for on-chip speed measurement
US20020140985A1 (en) 2001-04-02 2002-10-03 Hudson Kevin R. Color calibration for clustered printing
US20030086300A1 (en) 2001-04-06 2003-05-08 Gareth Noyes FPGA coprocessing system
US7930437B2 (en) 2001-04-11 2011-04-19 Mellanox Technologies Ltd. Network adapter with shared database for message context information
US7929539B2 (en) 2001-04-11 2011-04-19 Mellanox Technologies Ltd. Multiple queue pair access with a single doorbell
US20090213856A1 (en) 2001-05-04 2009-08-27 Slt Logic Llc System and Method for Providing Transformation of Multi-Protocol Packets in a Data Stream
US7688853B2 (en) 2001-05-04 2010-03-30 Slt Logic, Llc System and method for hierarchical policing of flows and subflows of a data stream
US7978606B2 (en) 2001-05-04 2011-07-12 Slt Logic, Llc System and method for policing multiple data flows and multi-protocol data flows
US7835375B2 (en) 2001-05-04 2010-11-16 Slt Logic, Llc Method and apparatus for providing multi-protocol, multi-stage, real-time frame classification
US6438130B1 (en) 2001-06-28 2002-08-20 Mellanox Technologies Ltd. Forwarding database cache
US20030007165A1 (en) 2001-07-05 2003-01-09 Hudson Kevin R. Direction-dependent color conversion in bidirectional printing
US7136397B2 (en) 2001-08-20 2006-11-14 Slt Logic Llc Network architecture and system for delivering bi-directional xDSL based services
US7151744B2 (en) 2001-09-21 2006-12-19 Slt Logic Llc Multi-service queuing method and apparatus that provides exhaustive arbitration, load balancing, and support for rapid port failover
US7099275B2 (en) 2001-09-21 2006-08-29 Slt Logic Llc Programmable multi-service queue scheduler
US20030058459A1 (en) 2001-09-26 2003-03-27 Yifeng Wu Generalized color calibration architecture and method
US6904534B2 (en) 2001-09-29 2005-06-07 Hewlett-Packard Development Company, L.P. Progressive CPU sleep state duty cycle to limit peak power of multiple computers on shared power distribution unit
US20030063299A1 (en) 2001-10-02 2003-04-03 Cowan Philip B. Color calibration method and apparatus
US20030065856A1 (en) 2001-10-03 2003-04-03 Mellanox Technologies Ltd. Network adapter with multiple event queues
US20030081060A1 (en) 2001-10-29 2003-05-01 Huanzhao Zeng Method and system for management of color through conversion between color spaces
US7240350B1 (en) 2002-01-07 2007-07-03 Slt Logic, Llc System and method for providing communications to processes
US7093158B2 (en) 2002-03-11 2006-08-15 Hewlett-Packard Development Company, L.P. Data redundancy in a hot pluggable, large symmetric multi-processor system
US20030172330A1 (en) 2002-03-11 2003-09-11 Barron Dwight L. Data redundancy in a hot pluggable, large symmetric multi-processor system
US7245627B2 (en) 2002-04-23 2007-07-17 Mellanox Technologies Ltd. Sharing a network interface card among multiple hosts
US20030202043A1 (en) 2002-04-24 2003-10-30 Huanzhao Zeng Determination of control points for construction of first color space-to-second color space look-up table
US6667918B2 (en) 2002-05-01 2003-12-23 Mellanox Technologies Ltd. Self-repair of embedded memory arrays
US7149227B2 (en) 2002-05-31 2006-12-12 Mellanox Technologies Ltd. Round-robin arbiter with low jitter
US20070220183A1 (en) 2002-07-23 2007-09-20 Mellanox Technologies Ltd. Receive Queue Descriptor Pool
US7143412B2 (en) 2002-07-25 2006-11-28 Hewlett-Packard Development Company, L.P. Method and apparatus for optimizing performance in a multi-processing system
US7631106B2 (en) 2002-08-15 2009-12-08 Mellanox Technologies Ltd. Prefetching of receive queue descriptors
US20080140574A1 (en) 2002-11-06 2008-06-12 Boucher Laurence B Enabling an enhanced function of an electronic device
US7386619B1 (en) 2003-01-06 2008-06-10 Slt Logic, Llc System and method for allocating communications to processors in a multiprocessor system
US7316017B1 (en) 2003-01-06 2008-01-01 Slt Logic, Llc System and method for allocatiing communications to processors and rescheduling processes in a multiprocessor system
US20080301406A1 (en) 2003-01-06 2008-12-04 Van Jacobson System and method for allocating communications to processors in a multiprocessor system
US20040141642A1 (en) 2003-01-22 2004-07-22 Huanzhao Zeng Color space conversion using interpolation
US20040156393A1 (en) 2003-02-12 2004-08-12 Silverback Systems, Inc. Architecture and API for of transport and upper layer protocol processing acceleration
US20040190533A1 (en) 2003-03-27 2004-09-30 Prashant Modi Method and apparatus for performing connection management with multiple stacks
US7103744B2 (en) 2003-03-27 2006-09-05 Hewlett-Packard Development Company, L.P. Binding a memory window to a queue pair
US20040193825A1 (en) 2003-03-27 2004-09-30 Garcia David J. Binding a memory window to a queue pair
US20040190557A1 (en) 2003-03-27 2004-09-30 Barron Dwight L. Signaling packet
US7554993B2 (en) 2003-03-27 2009-06-30 Hewlett-Packard Development Company, L.P. Method and apparatus for performing connection management with multiple stacks
US20040193734A1 (en) 2003-03-27 2004-09-30 Barron Dwight L. Atomic operations
US7502826B2 (en) 2003-03-27 2009-03-10 Hewlett-Packard Development Company, L.P. Atomic operations
US20040210754A1 (en) 2003-04-16 2004-10-21 Barron Dwight L. Shared security transform device, system and methods
US20070277036A1 (en) 2003-05-23 2007-11-29 Washington University, A Corporation Of The State Of Missouri Intelligent data storage and processing using fpga devices
US20040249881A1 (en) 2003-06-05 2004-12-09 Jha Ashutosh K. Transmitting commands and information between a TCP/IP stack and an offload unit
US20040249998A1 (en) 2003-06-05 2004-12-09 Anand Rajagopalan Uploading TCP frame data to user buffers and buffers in system memory
US20040252685A1 (en) 2003-06-13 2004-12-16 Mellanox Technologies Ltd. Channel adapter with integrated switch
US20050008223A1 (en) 2003-07-09 2005-01-13 Huanzhao Zeng Representing extended color gamut information
US20050018221A1 (en) 2003-07-21 2005-01-27 Huanzhao Zeng Conforming output intensities of pens
US20050039187A1 (en) 2003-08-12 2005-02-17 Avakian Arra E. Instrumenting java code by modifying bytecodes
US20050039172A1 (en) 2003-08-12 2005-02-17 Jeffrey Rees Synthesizing application response measurement (ARM) instrumentation
US20050039171A1 (en) 2003-08-12 2005-02-17 Avakian Arra E. Using interceptors and out-of-band data to monitor the performance of Java 2 enterprise edition (J2EE) applications
US20050038941A1 (en) 2003-08-14 2005-02-17 Mallikarjun Chadalapaka Method and apparatus for accessing a memory
US7617376B2 (en) 2003-08-14 2009-11-10 Hewlett-Packard Development Company, L.P. Method and apparatus for accessing a memory
US7757232B2 (en) 2003-08-14 2010-07-13 Hewlett-Packard Development Company, L.P. Method and apparatus for implementing work request lists
US20050038918A1 (en) 2003-08-14 2005-02-17 Hilland Jeffrey R. Method and apparatus for implementing work request lists
US20050066333A1 (en) 2003-09-18 2005-03-24 Krause Michael R. Method and apparatus for providing notification
US7404190B2 (en) 2003-09-18 2008-07-22 Hewlett-Packard Development Company, L.P. Method and apparatus for providing notification via multiple completion queue handlers
US7653754B2 (en) 2004-01-05 2010-01-26 Mellanox Technologies Ltd. Method, system and protocol that enable unrestricted user-level access to a network interface adapter
US20050172181A1 (en) 2004-01-16 2005-08-04 Mellanox Technologies Ltd. System and method for production testing of high speed communications receivers
US20050219278A1 (en) 2004-03-30 2005-10-06 Hudson Kevin R Image formation with a flexible number of passes
US20050219314A1 (en) 2004-03-30 2005-10-06 David Donovan Formation of images
US20050231751A1 (en) 2004-04-15 2005-10-20 Yifeng Wu Image processing system and method
US20060026443A1 (en) 2004-07-29 2006-02-02 Mcmahan Larry N Communication among partitioned devices
US7650386B2 (en) 2004-07-29 2010-01-19 Hewlett-Packard Development Company, L.P. Communication among partitioned devices
US7502870B1 (en) * 2004-08-13 2009-03-10 Sun Microsystems, Inc. Method for receiving network communication and apparatus for performing the same
US20060045098A1 (en) 2004-08-31 2006-03-02 Krause Michael R System for port mapping in a network
US7835380B1 (en) 2004-10-19 2010-11-16 Broadcom Corporation Multi-port network interface device with shared processing resources
US7403535B2 (en) 2004-12-14 2008-07-22 Hewlett-Packard Development Company, L.P. Aggregation of network resources providing offloaded connections between applications over a network
US20060126619A1 (en) 2004-12-14 2006-06-15 Teisberg Robert R Aggregation over multiple processing nodes of network resources each providing offloaded connections between applications over a network
US7551614B2 (en) 2004-12-14 2009-06-23 Hewlett-Packard Development Company, L.P. Aggregation over multiple processing nodes of network resources each providing offloaded connections between applications over a network
US20060165074A1 (en) 2004-12-14 2006-07-27 Prashant Modi Aggregation of network resources providing offloaded connections between applications over a network
US20060193318A1 (en) 2005-02-28 2006-08-31 Sriram Narasimhan Method and apparatus for processing inbound and outbound quanta of data
US20060228637A1 (en) 2005-04-11 2006-10-12 Jackson Bruce J Color filter
US20060248191A1 (en) 2005-04-27 2006-11-02 Hudson Charles L Aggregation of hybrid network resources operable to support both offloaded and non-offloaded connections
US7580415B2 (en) 2005-04-27 2009-08-25 Hewlett-Packard Development Company, L.P. Aggregation of hybrid network resources operable to support both offloaded and non-offloaded connections
US7580495B2 (en) 2005-06-30 2009-08-25 Slt Logic Llc Mixer-based phase control
US7573967B2 (en) 2005-07-01 2009-08-11 Slt Logic Llc Input threshold adjustment in a synchronous data sampling circuit
US20100138840A1 (en) 2005-08-23 2010-06-03 Mellanox Technologies Ltd. System and method for accelerating input/output access operation on a virtual machine
US7285996B2 (en) 2005-09-30 2007-10-23 Slt Logic, Llc Delay-locked loop
US7813460B2 (en) 2005-09-30 2010-10-12 Slt Logic, Llc High-speed data sampler with input threshold adjustment
US20100198850A1 (en) 2005-12-02 2010-08-05 Exegy Incorporated Method and Device for High Performance Regular Expression Pattern Matching
US7702629B2 (en) 2005-12-02 2010-04-20 Exegy Incorporated Method and device for high performance regular expression pattern matching
US7945528B2 (en) 2005-12-02 2011-05-17 Exegy Incorporated Method and device for high performance regular expression pattern matching
US20070188351A1 (en) 2005-12-16 2007-08-16 Andrew Brown Hardware enablement using an interface
US8286193B2 (en) 2006-01-10 2012-10-09 Solarflare Communications, Inc. Processing received data
US7827442B2 (en) 2006-01-23 2010-11-02 Slt Logic Llc Shelf management controller with hardware/software implemented dual redundant configuration
US7954114B2 (en) 2006-01-26 2011-05-31 Exegy Incorporated Firmware socket module for FPGA-based pipeline processing
US20070237327A1 (en) 2006-03-23 2007-10-11 Exegy Incorporated Method and System for High Throughput Blockwise Independent Encryption/Decryption
US7518164B2 (en) 2006-03-29 2009-04-14 Mellanox Technologies Ltd. Current-triggered low turn-on voltage SCR
US20100188140A1 (en) 2006-04-07 2010-07-29 Mellanox Technologies Ltd. Accurate Global Reference Voltage Distribution System With Local Reference Voltages Referred To Local Ground And Locally Supplied Voltage
US20070260602A1 (en) 2006-05-02 2007-11-08 Exegy Incorporated Method and Apparatus for Approximate Pattern Matching
US7636703B2 (en) 2006-05-02 2009-12-22 Exegy Incorporated Method and apparatus for approximate pattern matching
US20090302923A1 (en) 2006-05-08 2009-12-10 Mellanox Technologies Ltd. Terminated input buffer with offset cancellation circuit
US20110178918A1 (en) 2006-06-19 2011-07-21 Exegy Incorporated High Speed Processing of Financial Information Using FPGA Devices
US20110040701A1 (en) 2006-06-19 2011-02-17 Exegy Incorporated Method and System for High Speed Options Pricing
US20110184844A1 (en) 2006-06-19 2011-07-28 Exegy Incorporated High Speed Processing of Financial Information Using FPGA Devices
US20110178957A1 (en) 2006-06-19 2011-07-21 Exegy Incorporated High Speed Processing of Financial Information Using FPGA Devices
US20110178919A1 (en) 2006-06-19 2011-07-21 Exegy Incorporated High Speed Processing of Financial Information Using FPGA Devices
US20110178917A1 (en) 2006-06-19 2011-07-21 Exegy Incorporated High Speed Processing of Financial Information Using FPGA Devices
US20080008205A1 (en) 2006-07-07 2008-01-10 Jung Byung Kwon DATA ACCELERATION APPARATUS FOR iSCSI AND iSCSI STORAGE SYSTEM USING THE SAME
US20080024586A1 (en) 2006-07-31 2008-01-31 Rodolfo Jodra Barron System and method for calibrating a beam array of a printer
US20110004457A1 (en) 2006-08-07 2011-01-06 Voltaire Ltd. Service-oriented infrastructure management
US7801027B2 (en) 2006-08-30 2010-09-21 Mellanox Technologies Ltd. Auto-negotiation by nodes on an infiniband fabric
US20090201926A1 (en) 2006-08-30 2009-08-13 Mellanox Technologies Ltd Fibre channel processing by a host channel adapter
US7725556B1 (en) 2006-10-27 2010-05-25 Hewlett-Packard Development Company, L.P. Computer system with concurrent direct memory access
US20080148400A1 (en) 2006-10-31 2008-06-19 Hewlett-Packard Development Company, L.P. Method and apparatus for enforcement of software licence protection
US20080115217A1 (en) 2006-10-31 2008-05-15 Hewlett-Packard Development Company, L.P. Method and apparatus for protection of a computer system from malicious code attacks
US20080115216A1 (en) 2006-10-31 2008-05-15 Hewlett-Packard Development Company, L.P. Method and apparatus for removing homogeneity from execution environment of computing system
US20080109526A1 (en) 2006-11-06 2008-05-08 Viswanath Subramanian Rdma data to responder node coherency domain
US20080126509A1 (en) 2006-11-06 2008-05-29 Viswanath Subramanian Rdma qp simplex switchless connection
US20130007000A1 (en) 2006-11-13 2013-01-03 Exegy Incorporated Method and System for High Performance Integration, Processing and Searching of Structured and Unstructured Data Using Coprocessors
US8156101B2 (en) 2006-11-13 2012-04-10 Exegy Incorporated Method and system for high performance integration, processing and searching of structured and unstructured data using coprocessors
US20080135774A1 (en) 2006-12-08 2008-06-12 Asml Netherlands B.V. Scatterometer, a lithographic apparatus and a focus analysis method
US20080147828A1 (en) 2006-12-18 2008-06-19 Enstone Mark R Distributed Configuration Of Network Interface Cards
US7848322B2 (en) 2007-03-26 2010-12-07 Voltaire Ltd. Device, system and method of UDP communication
US7864787B2 (en) 2007-03-26 2011-01-04 Voltaire Ltd. Device, system and method of multicast communication
US7856488B2 (en) 2007-03-30 2010-12-21 Hewlett-Packard Development Company, L.P. Electronic device profile migration
US20080244060A1 (en) 2007-03-30 2008-10-02 Cripe Daniel N Electronic device profile migration
WO2008127672A2 (en) 2007-04-11 2008-10-23 Slt Logic Llc Modular blade for providing scalable mechanical, electrical and environmental functionality in the enterprise using advanced tca boards
US20080304519A1 (en) 2007-06-06 2008-12-11 Hewlett-Packard Development Company, L.P. Method for ethernet power savings on link aggregated groups
US7802071B2 (en) 2007-07-16 2010-09-21 Voltaire Ltd. Device, system, and method of publishing information to multiple subscribers
US20090060197A1 (en) 2007-08-31 2009-03-05 Exegy Incorporated Method and Apparatus for Hardware-Accelerated Encryption/Decryption
US20090165003A1 (en) 2007-12-21 2009-06-25 Van Jacobson System and method for allocating communications to processors and rescheduling processes in a multiprocessor system
US20090182683A1 (en) 2008-01-11 2009-07-16 Exegy Incorporated Method and System for Low Latency Basket Calculation
US20090183057A1 (en) 2008-01-15 2009-07-16 Neterion , Inc. OFFLOADING iSCSI WITHOUT TOE
US20110029669A1 (en) 2008-04-28 2011-02-03 Mike Chuang Adjustable Server-Transmission Rates Over Fixed-Speed Backplane Connections Within A Multi-Server Enclosure
WO2009134219A1 (en) 2008-04-28 2009-11-05 Hewlett-Packard Development Company, L.P. Adjustable server-transmission rates over fixed-speed backplane connections within a multi-server enclosure
US20090268612A1 (en) 2008-04-29 2009-10-29 Google Inc. Method and apparatus for a network queuing engine and congestion management gateway
WO2009136933A1 (en) 2008-05-08 2009-11-12 Hewlett-Packard Development Company, L.P. A method for interfacing a fibre channel network with an ethernet based network
US20110044344A1 (en) 2008-05-08 2011-02-24 Hudson Charles L A method for interfacing a fibre channel network with an ethernet based network
US20090287628A1 (en) 2008-05-15 2009-11-19 Exegy Incorporated Method and System for Accelerated Stream Processing
WO2010020907A2 (en) 2008-08-21 2010-02-25 Voltaire Ltd. Device, system, and method of distributing messages
US20100088437A1 (en) 2008-10-06 2010-04-08 Mellanox Technologies Ltd Infiniband adaptive congestion control adaptive marking rate
US7921178B2 (en) 2008-12-04 2011-04-05 Voltaire Ltd. Device, system, and method of accessing storage
US20120089496A1 (en) 2008-12-15 2012-04-12 Exegy Incorporated Method and Apparatus for High-Speed Processing of Financial Market Depth Data
US20120095893A1 (en) 2008-12-15 2012-04-19 Exegy Incorporated Method and apparatus for high-speed processing of financial market depth data
US20120089497A1 (en) 2008-12-15 2012-04-12 Exegy Incorporated Method and Apparatus for High-Speed Processing of Financial Market Depth Data
US20100169880A1 (en) 2008-12-25 2010-07-01 Voltaire Ltd. Virtual input-output connections for machine virtualization
US20100189206A1 (en) 2009-01-28 2010-07-29 Mellanox Technologies Ltd Precise Clock Synchronization
WO2010087826A1 (en) 2009-01-29 2010-08-05 Hewlett-Packard Development Company, L.P. Connecting ports of one or more electronic devices to different subsets of networks based on different operating modes
US20100265849A1 (en) 2009-04-21 2010-10-21 Voltaire Ltd Spanning tree root selection in a hierarchical network
US8000336B2 (en) 2009-04-21 2011-08-16 Voltaire Ltd. Spanning tree root selection in a hierarchical network
US20100274876A1 (en) 2009-04-28 2010-10-28 Mellanox Technologies Ltd Network interface device with memory management capabilities
US20110010557A1 (en) 2009-07-07 2011-01-13 Mellanox Technologies Ltd Control message signature for device control
US20110029847A1 (en) 2009-07-30 2011-02-03 Mellanox Technologies Ltd Processing of data integrity field
US20110058571A1 (en) 2009-09-09 2011-03-10 Mellanox Technologies Ltd. Data switch with shared port buffers
US20110083064A1 (en) 2009-10-04 2011-04-07 Mellanox Technologies Ltd Processing of block and transaction signatures
US7934959B2 (en) 2009-10-06 2011-05-03 Mellanox Technologies Ltd. Adapter for pluggable module
WO2011043769A1 (en) 2009-10-07 2011-04-14 Hewlett-Packard Development Company, L.P. Notification protocol based endpoint caching of host memory
US20110096668A1 (en) 2009-10-26 2011-04-28 Mellanox Technologies Ltd. High-performance adaptive routing
WO2011053305A1 (en) 2009-10-30 2011-05-05 Hewlett-Packard Development Co Thermal bus bar for a blade enclosure
WO2011053330A1 (en) 2009-11-02 2011-05-05 Hewlett-Packard Development Company, L.P. Multiprocessing computing with distributed embedded switching
US20110113083A1 (en) 2009-11-11 2011-05-12 Voltaire Ltd Topology-Aware Fabric-Based Offloading of Collective Functions
US20110119673A1 (en) 2009-11-15 2011-05-19 Mellanox Technologies Ltd. Cross-channel network operation offloading for collective operations
US20110116512A1 (en) 2009-11-19 2011-05-19 Mellanox Technologies Ltd Dynamically-Connected Transport Service
US20110173352A1 (en) 2010-01-13 2011-07-14 Mellanox Technologies Ltd Power Reduction on Idle Communication Lanes
US8326816B2 (en) 2010-05-06 2012-12-04 Oracle International Corporation Fine grain synchronization for database replay
US20120102245A1 (en) 2010-10-20 2012-04-26 Gole Abhijeet P Unified i/o adapter
EP2463782A2 (en) 2010-12-09 2012-06-13 Solarflare Communications Inc Encapsulated Accelerator
US20120246052A1 (en) 2010-12-09 2012-09-27 Exegy Incorporated Method and Apparatus for Managing Orders in Financial Markets

Non-Patent Citations (162)

* Cited by examiner, † Cited by third party
Title
"MPI: A Message-Passing Interface Standard," various forum members, Message-Passing Interface Forum, University of Tennessee, Knoxville, May 5, 1994 (corrected version), 239 pp.
"NVIDIA GPUDirect™ Technology—Accelerating GPU-based Systems," Mellanox Tech. Brief, May 2010, 2pp: <http://www.mellanox.com/pdf/whitepapers/TB—GPU—Direct.pdf>.
"NVIDIA Tesla GPUs to Communicate Faster Over Mellanox InfiniBand Networks," press release dated Nov. 25, 2009, Portland OR, 3 pp: <http://gpgpu.org/2009/11/25/nvidia-tesla-mellanox-infiniband>.
A. Edwards, et al.; ACM Computer Communication Review, vol. 24, No. 4, pp. 14-23, Oct. 1994.
A. Edwards, S. Muir, "Experiences implementing a high performance TCP in user-space," ACM Computer Communication Review, vol. 25, No. 4, Oct. 1995, pp. 196-205.
A. Romanow and S. Floyd; ACM Computer Communication Review, vol. 24, No. 4, p. 79-88, Oct. 1994.
Andrew D. Birrell, et al.; Communications of the ACM, vol. 25, Issue 4, pp. 260-274, Apr. 1982.
Andy Currid, "TCP Offload to the Rescue," ACM Queue, vol. 2, No. 3, May 1, 2004, pp. 58-65.
B. Leslie, et al., "User-level Device Drivers: Achieved Performance," J. Comput. Sci. & Technol., vol. 20, Sep. 5, 2005, 17pp.
Babak Falsafi, et al.; Proceedings of the 1994 conference on Supercomputing, pp. 380-389, Washington D.C., Nov. 14, 1994.
Bilic Hrvoye, et al.; article in Proceedings of the 9th Symposium on High Performance Interconnects, "Deferred Segmentation for Wire-Speed Transmission of Large TCP Frames over Standard GbE Networks," Aug. 22, 2001, 5pp.
Bilic Hrvoye, et al.; presentation slides from 9th Symposium on High Performance Interconnects, "Deferred Segmentation for Wire-Speed Transmission of Large TCP Frames over Standard GbE Networks," Aug. 22, 2001, 9pp.
Bruce Lowekamp, et al., "Topology discovery for large ethernet networks," ACM SIGCOMM Computer Communication Review-Proc. 2001 SIGCOMM conf., vol. 31, No. 4, Oct. 1, 2001, pp. 237-248.
Bruce Lowekamp, et al., "Topology discovery for large ethernet networks," ACM SIGCOMM Computer Communication Review—Proc. 2001 SIGCOMM conf., vol. 31, No. 4, Oct. 1, 2001, pp. 237-248.
Bruce S. Davie, "A host-network interface architecture for ATM," ACM Computer Communication Review, vol. 21, No. 4, Sep. 1991, pp. 307-315.
C. A. Thekkath, et al., "Implementing network protocols at user level," ACM Computer Communication Review, vol. 23, No. 4, Oct. 1993, pp. 64-73.
C. Brendan S. Traw, et al.; ACM Computer Communication Review, vol. 21, No. 4, p. 317-325, Sep. 1991.
C. Kline, "Supercomputers on the internet: a case study," ACM Computer Communication Review, vol. 17, No. 5, Aug. 1987, pp. 27-33.
C. Partridge, J. Hughes, J. Stone; ACM Computer Communication Review, vol. 25, No. 4, p. 68-76, Oct. 1995.
C. Traw and J. Smith; IEEE Journal on Selected Areas in Communications, pp. 240-253, Feb. 1993.
Charles Kalmanek; ACM Computer Communication Review, vol. 32, No. 5, pp. 13-19, Nov. 2, 2002.
Charles P. Thacker and Lawrence C. Stewart; ACM Operating Systems Review, vol. 21, Issue 4, p. 164-172, 1987, Oct. 1997.
Cheng Jin, et al.; Proceedings of IEEE Infocom 2004, pp. 1246-1259, Mar. 7, 2004.
Chi-Chao Chang, et al., "Low-Latency Communication on the IBM RISC System/6000 SP," Proc. 1996 ACM/IEEE conference on Supercomputing, Pittsburgh, Nov. 17, 1996, 17pp.
Chris Maeda, Brian Bershad; ACM Operating Systems Review, vol. 27, Issue 5, p. 244-255, Dec. 1993.
Christopher A. Kent, Jeffrey C. Mogul; ACM Computer Communication Review, vol. 17, No. 5, pp. 390-401, Oct. 1987.
Craig Partridge; ACM Computer Communication Review, vol. 20, No. 1, p. 44-53, Jan. 1990.
D. D. Clark and D. L. Tennenhouse; ACM Computer Communication Review, vol. 20, No. 4, pp. 200-208, Sep. 1990.
D. L. Tennenhouse, D. J. Wetherall; ACM Computer Communication Review, vol. 26, No. 2, pp. 15-20, Apr. 1996.
Danny Cohen, et al.; ACM Computer Communication Review, vol. 23, No. 4, p. 32-44, Jul. 1993.
Danny Cohen, Gregory Finn, Robert Felderman, Annette DeSchon, "Atomic: A High Speed Local Communication Architecture," Journal of High Speed Networks, vol. 3:1 (1994), pp. 1-28.
Danny Cohen, Gregory Finn, Robert Felderman, Annette DeSchon; Made available by authors, Jan. 10, 1992, "Atomic: A Local Communication Network Created through Repeated Application of Multicomputing Components," 11 pp + cover.
David A. Borman; ACM Computer Communication Review, vol. 19, No. 2, p. 11-15, Apr. 1989.
David D. Clark, et al.; IEEE Communications Magazine, vol. 27, No. 6, pp. 23-29, Jun. 1989.
David D. Clark; ACM Computer Communication Review, vol. 18, No. 4, pp. 106-114, Aug. 1988.
David R. Boggs, et al.; ACM Computer Communication Review, vol. 18, No. 4, p. 222-234, Aug. 1988.
David R. Cheriton; ACM Computer Communication Review, vol. 19, No. 4, p. 158-169, Sep. 1989.
David Wetherall; ACM Computer Communication Review, vol. 36, No. 3, pp. 77-78, Jul. 6, 2006.
Declaration of Dr. Gregory L. Chesson in Support of Microsoft's Opposition to Alacritech's Motion for Preliminary Injunction; United States District Court, Northern District California, San Francisco Division, Case No. 3:04-cv-03284-JSW, filed Feb. 4, 2005, 289 pp including exhibits (declaration is 25 pp).
Derek Chiou, Boon S. Ang, et al., "StarT-Voyager: A Flexible Platform for Exploring Scalable SMP Issues," Proc. 1998 ACM/IEEE conference on Supercomputing, Orlando, Florida, Nov. 7, 1998, 20pp.
Derek McAuley, Rolf Neugebauer, "A case for virtual channel processors," NICELI '03 Proc.of the ACM SIGCOMM workshop, Aug. 2003, pp. 237-242.
Derek Robert McAuley, "Protocol Design for High Speed Networks," Ph.D. Thesis, University of Cambridge, Sep. 1989, 104 pp.
Dickman, L., "Protocol OffLoading vs OnLoading in High Performance Networks," 14th Symposium on High Performance Interconnects, Aug. 23, 2006, 8pp.
E. Blanton and M. Allman, "On Making TCP More Robust to Packet Reordering," ACM Computer Communication Review, vol. 32, No. 1, Jan. 2, 2002, pp. 20-30.
E. Ruetsche, "The architecture of a Gb/s multimedia protocol adapter," ACM Computer Communication Review, vol. 23, No. 3, Jul. 1993, pp. 59-68.
Ed Anderson, et al.; Proceedings of the 1997 ACM/IEEE conference on Supercomputing, p. 1-17, San Jose, California, Nov. 16, 1997.
Edward D. Lazowska, David A. Patterson, "Computing Research: A Looming Crisis," ACM Computer Communication Review, vol. 35, No. 2, Jul. 5, 2005, pp. 65-68.
EP 12185546.4—Extended European Search Report dated Jul. 13, 2013, 6 pages.
EP 13153148.5—1953—Extended European Search Report dated Feb. 19, 2014, 6 pages.
EP 13187725.0—1953—Extended European Search Report dated Feb. 19, 2014, (6 pages).
Eric C. Cooper, et al.; ACM Computer Communication Review, vol. 20, No. 4, p. 135-144, Sep. 1990.
Erich Ruetsche and Matthias Kaiserswerth, "TCP/IP on the parallel protocol engine," Proceedings of the IFIP TC6/WG6, 4 Fourth International Conference on High Performance Networking IV, Dec. 14, 1992, pp. 119-134.
F.F. Kuo, "The Aloha system," ACM Computer Communication Review, vol. 4 No. 1, Jan. 1974, pp. 5-8.
Gary S. Delp, et al.; ACM Computer Communication Review, vol. 18, No. 4, p. 165-174, Aug. 1988.
Gene Tsudik; ACM Computer Communication Review, vol. 22, No. 5, pp. 29-38, Oct. 1992.
Geoffray P., "Protocol off-loading vs on-loading in high-performance networks," 14th Symposium on High Performance Interconnects, Aug. 23, 2006, 5pp.
Gordon E. Moore; Electronics, vol. 38, No. 8, pp. 114-117, Apr. 19, 1965.
Greg Chesson, "The Evolution of XTP," Proceedings of the Third International Conference on High Speed Networking, Nov. 1991, 10 pp.
Greg Minshall, et al.; ACM Computer Communication Review, vol. 36, No. 3, pp. 79-92, Jul. 6, 2006.
Greg Regnier, et al.; Computer, IEEE Computer Society, vol. 37, No. 11, pp. 48-58, Nov. 4, 2004.
Greg Regnier, et al.; IEEE Micro, vol. 24, No. 1, p. 24-31, Jan. 1994.
Gregory G. Finn and Paul Mockapetris, "Netstation Architecture Multi-Gigabit Workstation Network Fabric," Proceedings of InterOp '94, Las Vegas, Nevada, May 1994, 9pp.
Gregory G. Finn; ACM Computer Communication Review, vol. 21, No. 5, p. 18-29, Oct. 1991.
H. K. Jerry Chu, "Zero-Copy TCP in Solaris," Proc. USENIX Annual Technical Conference, Jan. 1996, 13pp.
H. Kanakia and D. Cheriton; ACM Computer Communication Review, vol. 18, No. 4, p. 175-187, Aug. 1988.
Harvey J. Wassermann, et al.; Proceedings of the 1997 ACM/IEEE conference on Supercomputing, p. 1-11, San Jose, California, Nov. 16, 1997.
Hodges, S.J. et al., "Remoting Peripherals using Memory-Mapped Networks," Olivetti and Oracle Research Laboratory, Cambridge Univ., 1998, 3 pp: <http://www.cl.cam.ac.uk/research/dtg/www/publications/public/files/tr.98.6.pdf>.
Humaira Kamal, et al., "SCTP versus TCP for MPI," Proc.2005 ACM/IEEE conference on Supercomputing, Seattle, Washington, Nov. 12, 2005, 14pp.
Ian Leslie and Derek R. McAuley; ACM Computer Communication Review, vol. 21, No. 4, p. 327, Sep. 1991.
Ian M. Leslie, et al.; ACM Computer Communication Review, vol. 14, No. 2, pp. 2-9, Jun. 1984.
Ian Pratt and Keir Fraser; Proceedings of IEEE Infocom 2001, pp. 67-76, Apr. 22, 2001.
J. C. Mogul, "The case for persistent-connection HTTP," ACM Computer Communication Review, vol. 25, No. 4, Oct. 1995, pp. 299-313.
J. Carver Hill; Communications of the ACM, vol. 16, No. 6, p. 350-351, Jun. 1973.
J. Evans and T. Buller, "The end of history," IEEE Gigabit Networking Workshop, INFOCOM 2001, Anchorage, Alaska, Apr. 22, 2001.
J. Vis; ACM Computer Communication Review, vol. 24, No. 1, pp. 7-11, Jan. 1994.
Jack B. Dennis and Earl C. Van Horn; Communications of the ACM, vol. 9, No. 3, pp. 143-155, Mar. 1966.
Jeffrey C. Mogul; Proceedings of HotOS IX: The 9th Workshop on Hot Topics in Operating Systems, pp. 25-30, May 18, 2003.
Jeffrey R. Michel, "The Design and Evaluation of an Off-Host Communications Protocol Architecture," M.Sci. Thesis, University of Virginia, Aug. 1993, 139 pp.
Jenwei Hsieh, et al., "Architectural and Performance Evaluation of GigaNet and Myrinet Interconnects on Clusters of Small-Scale SMP Servers," Proc. 2000 ACM/IEEE conference on Supercomputing, Dallas, Texas, Nov. 4, 2000, 9pp.
Jiuxing Liu, et al., "Performance Comparison of MPI Implementations over InfiniBand, Myrinet and Quadrics," Proc. 2003 ACM/IEEE conference on Supercomputing, Phoenix, Arizona, Nov. 15, 2003, 14 pp. 58-71.
Joe Touch, et al.; "Atomic-2" slides, Gigabit Networking Workshop '97 Meeting, Kobe, Japan, Apr. 1997, 10pp.
Joe Touch, et al.; "Host-based Routing Using Peer DMA," Gigabit Networking Workshop '97 Meeting, Kobe, Japan, Apr. 1997, 2pp.
John M. McQuillan, et al.; Proceedings of the 6th Data Communications Symposium, p. 63, Nov. 1979.
John Nagle; ACM Computer Communication Review, vol. 14, No. 4, p. 11-17, Oct. 1984.
John Salmon, et al., "Scaling of Beowulf-class Distributed Systems," Proc. 1998 ACM/IEEE conference on Supercomputing, Orlando, Florida, Nov. 7, 1998, 18pp.
Jon Crowcroft, Derek McAuley, "ATM: A Retrospective on Systems Legacy or ‘A technology with a fabulous future behind it?’", ACM Computer Communication Review, vol. 32, No. 5, Nov. 2, 2002, pp. 11-12. (introduces 4 articles spanning pp. 13-27; by Kalmanek pp. 13-19, Grossman pp. 21-24, Crosby et al. pp. 25-28, and Smith pp. 29-37).
Jon Crowcroft, Derek McAuley, "ATM: A Retrospective on Systems Legacy or 'A technology with a fabulous future behind it?'", ACM Computer Communication Review, vol. 32, No. 5, Nov. 2, 2002, pp. 11-12. (introduces 4 articles spanning pp. 13-27; by Kalmanek pp. 13-19, Grossman pp. 21-24, Crosby et al. pp. 25-28, and Smith pp. 29-37).
Jon Crowcroft; ACM Computer Communication Review, vol. 36, No. 2, pp. 51-52, Apr. 6, 2006.
Jonathan Kay and Joseph Pasquale; ACM Computer Communication Review, vol. 23, No. 4, pp. 259-268, Oct. 1993.
Jonathan M. Smith and C. Brendan S. Traw; IEEE Network, vol. 7, Issue 4, pp. 44-52, Jul. 1993.
Jonathan Smith; ACM Computer Communication Review, vol. 32, No. 5, pp. 29-37, Nov. 2, 2002.
Jonathan Stone, Craig Partridge; ACM Computer Communication Review, vol. 30, No. 4, pp. 309-319, Oct. 2000.
Jose Carlos Sancho, et al., "Quantifying the Potential Benefit of Overlapping Communication and Computation in Large-Scale Scientific Applications," Proc.2006 ACM/IEEE conference on Supercomputing, Tampa, Florida, Nov. 11, 2006, 16pp.
Justin Hurwitz, Wu-chun Feng, "Initial end-to-end performance evaluation of 10-Gigabit Ethernet," Proc. 11th Symposium on High Performance Interconnects, Aug. 20, 2003, pp. 116-121.
K. Kleinpaste, P. Steenkiste, B. Zill; ACM Computer Communication Review, vol. 25, No. 4, p. 87-98, Oct. 1995.
Ken Calvert; ACM Computer Communication Review, vol. 36, No. 2, pp. 27-30, Apr. 6, 2006.
Kieran Mansley, et al.; Euro-Par Conference 2007, pp. 224-233, Rennes, France, Aug. 28, 2007.
L.S. Brakmo, et al., "TCP Vegas: new techniques for congestion detection and avoidance," ACM Computer Communication Review, vol. 24, No. 4, Oct. 1994, p. 24-35.
M. Allman, "TCP Byte Counting Refinements," ACM Computer Communication Review, vol. 29, No. 3, Jul. 1999, pp. 14-22.
M. de Vivo, et al.; ACM Computer Communication Review, vol. 29, No. 1, pp. 81-85, Jan. 1999.
M. Kaiserswerth; IEEE/ACM Transactions in Networking vol. 1, Issue 6, pp. 650-663, Dec. 1993.
M.V. Wilkes and R.M. Needham, "The Cambridge Model Distributed System," ACM SIGOPS Operating Systems Review, vol. 14, Issue 1, pp. 21-29, Jan. 1980.
Margaret L. Simmons and Harvey J. Wasserman; Proceedings of the 1988 ACM/IEEE conference on Supercomputing, p. 288-295, Orlando, Florida, Nov. 12, 1988.
Mark David Hayter, "A Workstation Architecture to Support Multimedia," Ph.D. Thesis, University of Cambridge, Sep. 1993, 111 pp.
Mark Hayter, Derek McAuley; ACM Operating Systems Review, vol. 25, Issue 4, p. 14-21, Oct. 1991.
Marvin Zelkowitz; Communications of the ACM, vol. 14, No. 6, p. 417-418, Jun. 1971.
Matthias Kaiserswerth; IEEE/ACM Transactions on Networking, vol. 1, No. 6, p. 650-663, Dec. 1993.
Mengjou Lin, et al., "Performance of High-Speed Network I/O Subsystems: Case Study of a Fibre Channel Network," Proc.1994 Conf. on Supercomputing, Washington D.C., Nov. 14, 1994, pp. 174-183.
Michael J. Dixon, "System support for multi-service traffic," University of Cambridge Computer Laboratory Technical Report No. 245, Jan. 1992, 113 pp.
Michael S. Warren, et al., "Avalon: An Alpha/Linux Cluster Achieves 10 Gflops for $150k," Proc. 1998 ACM/IEEE conference on Supercomputing, Orlando, Florida, Nov. 7, 1998, 10pp.
Mogul J., "TCP offload is a dumb idea whose time has come," USENIX Assoc., Proceedings of HotOS IX: The 9th Workshop on Hot Topics in Operating Systems, May 2003, pp. 24-30.
Montry G., OpenFabrics Alliance presentation slides, 14th Symposium on High Performance Interconnects, Aug. 23, 2006, 8pp.
Murali Rangarajan, et al., "TCP Servers: Offloading TCP Processing in Internet Servers. Design, Implementation, and Performance," Technical Report DCR-TR-481, Computer Science Department, Rutgers University, Mar. 2, 2002, 14 pp.
Nanette J. Boden, et al.; Draft of paper published in IEEE Micro, vol. 15, No. 1, pp. 29-36, 1995, Nov. 15, 1994.
NR Adiga, et al., "An Overview of the BlueGene/L Supercomputer," Proceedings of the 2002 ACM/IEEE conference on Supercomputing, Baltimore, Nov. 16, 2002, pp. 1-22.
O. Angin, et al.; ACM Computer Communication Review, vol. 27, No. 3, pp. 100-117, Jul. 1997.
P. Balaji, et al., "Head-to TOE Evaluation of High-Performance Sockets over Protocol Offload Entines," Proceedings of the IEEE International Conference on Cluster Computing, Sep. 5, 2005, 10pp.
P. Druschel, et al. "Experiences with a high-speed network adaptor: a software perspective," ACM Computer Communication Review, vol. 24, No. 4, Oct. 1994, pp. 2-13.
P. Kermani and L. Kleinrock; Computer Networks, vol. 3, No. 4, pp. 267-286, Sep. 1979.
Parry Husbands and James C. Hoe; Proceedings of the 1998 ACM/IEEE conference on Supercomputing, p. 1-15, Orlando, Florida, Nov. 7, 1998.
Pasi Sarolahti, et al., "F-RTO: an enhanced recovery algorithm for TCP retransmission timeouts," ACM Computer Communication Review, vol. 33, No. 2, Apr. 3, 2003, pp. 51-63.
Patrick Crowley, et al.; Proceedings of the 14th international conference on Supercomputing, pp. 54-65, Santa Fe, New Mexico, May 8, 2000.
Patrick Geoffray, "A Critique of RDMA," HPCWire article, Aug. 18, 2006, 7pp. Available at http://www.hpcwire.com/hpcwire/2006-08-18/a-critique-of-rdma-1.html.
Patrick Geoffray, "A Critique of RDMA," HPCWire article, Aug. 18, 2006, 7pp. Available at http://www.hpcwire.com/hpcwire/2006-08-18/a—critique—of—rdma-1.html.
Paul E. McKenney and Ken F. Dove, "Efficient demultiplexing of incoming TCP packets," ACM Computer Communication Review, vol. 22, No. 4, Oct. 1992, pp. 269-279.
Paul Ronald Barham, "Devices in a Multi-Service Operating System," Ph.D. thesis, University of Cambridge, Jul. 1996, 142pp.
Paul V. Mockapetris, Kevin J. Dunlap; ACM Computer Communication Review, vol. 18, No. 4, pp. 123-133, Aug. 1988.
Peter Druschel and Larry L. Peterson; ACM Operating Systems Review, vol. 27, Issue 5, p. 189-202, Dec. 1993.
Peter Steenkiste, "Analyzing communication latency using the Nectar communication processor," ACM Computer Communication Review, vol. 22, No. 4, Oct. 1992, pp. 199-209.
Petrini F., "Protocol Off-loading vs On-loading in High-Performance Networks," 14th Symposium on High Performance Interconnects, Aug. 23, 2006, 4pp.
Philip Buonadonna, et al.; Proceedings of the 1998 ACM/IEEE conference on Supercomputing, p. 1-15, Orlando, Florida, Nov. 7, 1998.
Piyush Shivam, et al., "EMP: Zero-copy OS-bypass NIC-driven Gigabit Ethernet Message Passing," Proc. 2001 ACM/IEEE Conf. on Supercomputing, Denver, Nov. 10, 2001 (8 pp).
Pope, S.L. et al., "Enhancing Distributed Systems with Low-Latency Netowrking," Olivetti and Oracle Research Laboratory, Cambridge Univ. May 1998, 12 pp: <http://www.cl.cam.ac.uk/research/dtg/www/publications/public/files/tr.98.7.pdf>.
R. Braden, et al.; ACM Computer Communication Review, vol. 19, No. 2, p. 86-94, Apr. 1989.
R. Bush, D. Meyer; IETF Network Working Group, Request for Comments memo: 3439 [extending RFC1958], Dec. 2002, Some Internet Architectural Guidelines and Philosophy, 25 pp, The Internet Society (c)2002.
R. J. Black, I. Leslie, and D. McAuley; ACM Computer Communication Review, vol. 24, No. 4, p. 158-167, Oct. 1994.
Raj K. Singh, et al.; ACM Computer Communication Review, vol. 24, No. 3, p. 8-17, Jul. 1994.
Raj K. Singh, et al.; Proceedings of the 1993 ACM/IEEE conference on Supercomputing, p. 452-461, Portland, Oregon, Nov. 15, 1993.
Regnier G., "Protocol Onload vs. Offload," 14th Symposium on High Performance Interconnects, Aug. 23, 2006, 1pp.
Robert K. Brandriff, et al.; "Development of a TCP/IP for the IBM/370," ACM SIGCOMM Computer Communication Review, vol. 15, No. 4, Sep. 1985, pp. 2-8.
Robert M. Metcalfe and David R. Boggs; Communications of the ACM, vol. 19, Issue 7, pp. 395-404, Jul. 1976.
Robert Ross, et al., "A Case Study in Application I/O on Linux Clusters," Proc. 2001 ACM/IEEE Conf. on Supercomputing, Denver, Nov. 10, 2001, 17pp.
S. L. Pope, et al., "Enhancing Distributed Systems with Low-Latency Networking," Parallel and Distributed Computing and Networks, Brisbane, Australia, Dec. 1998, 10pp.
Sally Floyd; ACM Computer Communication Review, vol. 24, No. 5, p. 8-23, Oct. 1994.
Sayantan Sur, et al., "High-Performance and Scalable MPI over InfiniBand with Reduced Memory Usage; An In-Depth Performance Analysis," Proc. 2006 ACM/IEEE conference on Supercomputing, Tampa, Florida, Nov. 11, 2006, 13pp.
Srihari Makineni and Ravi Iyer; Proceedings of the 10th International Symposium on High Performance Computer Architecture, pp. 152, Feb. 14, 2004.
Steve Muir and Jonathan Smith, "Piglet: A Low-Intrusion Vertical Operating System," Technical Report MS-CIS-00-04, University of Pennsylvania, Jan. 2000, 15pp.
Steven J. Sistare, Christopher J. Jackson; Proceedings of the 2002 ACM/IEEE conference on Supercomputing, p. 1-15, Baltimore, Nov. 16, 2002.
Steven Pope, David Riddoch; ACM Computer Communication Review, vol. 37, No. 2, pp. 89-92, 2007, Mar. 19, 2007.
Stuart Wray, et al.; Proceedings of the International Conference on Multimedia Computing and Systems, p. 265-273, Boston, May 1994.
Sumitha Bhandarkar, et al., "LTCP: Improving the Permormance of TCP in Highspeed Networks," ACM Computer Communication Review, vol. 36, No. 1, Jan. 6, 2006, pp. 41-50.
Thomas Sterling, et al.; Proceedings of the 24th International Conference on Parallel Processing, pp. 11-14, Aug. 1995.
Thorsten von Eicken, et al.; ACM Operating Systems Review, vol. 29, Issue 5, p. 109-126, Dec. 1995.
Tom Kelly; ACM Computer Communication Review, vol. 33, No. 2, pp. 83-91, Apr. 3, 2003.
V. Cerf, et al.; ACM Computer Communication Review, vol. 6 No. 1, p. 1-18, Jan. 1976.
V. Jacobson; ACM Computer Communication Review, vol. 18, No. 4, p. 314-329, Aug. 1988.
Vinay Aggarwal, et al., summary, NICELI workshop, ACM SIGCOMM Computer Communication Review, vol. 33, No. 5, Oct. 3, 2003, pp. 75-80.
Vinton Cerf, Robert Kahn; IEEE Transactions on Communications, vol. COM-22, No. 5, pp. 637-648, May 1974.
W. E. Leland, et al.; ACM Computer Communication Review, vol. 23, No. 4, p. 85-95, Oct. 1993.
W. Feng and P. Tinnakornsrisuphap, "The Failure of TCP in High-Performance Computational Grids," Proc. 2000 ACM/IEEE conference on Supercomputing, Dallas, Texas, Nov. 4, 2000, 11p.
W. Feng, et al., "Permormance characterization of a 10-Gigabit Ethernet TOE," Proc. 13th Symp. on High Performance Interconnects, Aug. 17, 2005, pp. 58-63.
Wu-chun Feng, et al., "Optimizing 10-Gigabit Ethernet for Networks of Workstations, Clusters, and Grids: A Case Study," Proceedings of the 2003 ACM/IEEE conference on Supercomputing, Phoenix, Arizona, Nov. 15, 2003, 13 pp.

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9880964B2 (en) 2010-12-09 2018-01-30 Solarflare Communications, Inc. Encapsulated accelerator
US9892082B2 (en) 2010-12-09 2018-02-13 Solarflare Communications Inc. Encapsulated accelerator

Also Published As

Publication number Publication date
US20130080651A1 (en) 2013-03-28
EP2574000A3 (en) 2013-08-14
EP2574000A2 (en) 2013-03-27

Similar Documents

Publication Publication Date Title
US8458280B2 (en) Apparatus and method for packet transmission over a high speed network supporting remote direct memory access operations
US9043671B2 (en) Data protocol
US8213315B2 (en) Dynamically-connected transport service
US8699521B2 (en) Apparatus and method for in-line insertion and removal of markers
US20040049601A1 (en) Split socket send queue apparatus and method with efficient queue flow control, retransmission and sack support mechanisms
KR100974045B1 (en) Increasing tcp re-transmission process speed
CN101102305B (en) Method and system for managing network information processing
US7420931B2 (en) Using TCP/IP offload to accelerate packet filtering
US7171484B1 (en) Reliable datagram transport service
US8175116B2 (en) Multiprocessor system for aggregation or concatenation of packets
US7782905B2 (en) Apparatus and method for stateless CRC calculation
US7346707B1 (en) Arrangement in an infiniband channel adapter for sharing memory space for work queue entries using multiply-linked lists
US7475167B2 (en) Offloading data path functions
US7483376B2 (en) Method and apparatus for discovering path maximum transmission unit (PMTU)
US7580406B2 (en) Remote direct memory access segment generation by a network controller
US20070208820A1 (en) Apparatus and method for out-of-order placement and in-order completion reporting of remote direct memory access operations
EP2552080A2 (en) Chimney onload implementation of network protocol stack
US20100023626A1 (en) Method and apparatus for reducing host overhead in a socket server implementation
US8244906B2 (en) Method and system for transparent TCP offload (TTO) with a user space library
US7013419B2 (en) Reliable message transmission with packet-level resend
US7609696B2 (en) Storing and accessing TCP connection information
EP1732010A1 (en) A system-on-a-chip (SoC) device with integrated support for ethernet, TCP, ISCSI, RDMA, and network application acceleration
US8176187B2 (en) Method, system, and program for enabling communication between nodes
US7535907B2 (en) TCP engine
US8006169B2 (en) Data transfer error checking

Legal Events

Date Code Title Description
AS Assignment

Owner name: SOLARFLARE COMMUNICATIONS, INC., CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:POPE, STEVEN L.;RIDDOCH, DAVID J.;SIGNING DATES FROM 20121002 TO 20121004;REEL/FRAME:029151/0285

AS Assignment

Owner name: SOLARFLARE COMMUNICATIONS INC., CALIFORNIA

Free format text: CHANGE OF ADDRESS OF THE ASSIGNEE;ASSIGNOR:SOLARFLARE COMMUNICATIONS INC.;REEL/FRAME:029636/0295

Effective date: 20130104

STCF Information on status: patent grant

Free format text: PATENTED CASE

AS Assignment

Owner name: PACIFIC WESTERN BANK, NORTH CAROLINA

Free format text: SECURITY INTEREST;ASSIGNOR:SOLARFLARE COMMUNICATIONS, INC.;REEL/FRAME:038363/0589

Effective date: 20160331

AS Assignment

Owner name: ALLY BANK, NEW YORK

Free format text: SECURITY INTEREST;ASSIGNOR:SOLARFLARE COMMUNICATIONS, INC.;REEL/FRAME:041902/0527

Effective date: 20170223

AS Assignment

Owner name: SOLARFLARE COMMUNICATIONS, INC., CALIFORNIA

Free format text: RELEASE BY SECURED PARTY;ASSIGNOR:PACIFIC WESTERN BANK;REEL/FRAME:042820/0890

Effective date: 20170615

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YR, SMALL ENTITY (ORIGINAL EVENT CODE: M2551); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY

Year of fee payment: 4